Abstract
Soil microorganisms often thrive as microcolonies or biofilms within pores of soil aggregates exposed to the soil atmosphere. However, previous studies on the physiology of soil ammonia-oxidizing microorganisms (AOMs), which play a critical role in the nitrogen cycle, were primarily conducted using freely suspended AOM cells (planktonic cells) in liquid media. In this study, we examined the growth of two representative soil ammonia-oxidizing archaea (AOA), Nitrososphaera viennensis EN76 and “Nitrosotenuis chungbukensis” MY2, and a soil ammonia-oxidizing bacterium, Nitrosomonas europaea ATCC 19718 on polycarbonate membrane filters floated on liquid media to observe their adaptation to air-exposed solid surfaces. Interestingly, ammonia oxidation activities of N. viennensis EN76 and “N. chungbukensis” MY2 were significantly repressed on floating filters compared to the freely suspended cells in liquid media. Conversely, the ammonia oxidation activity of N. europaea ATCC 19718 was comparable on floating filters and liquid media. N. viennensis EN76 and N. europaea ATCC 19718 developed microcolonies on floating filters. Transcriptome analysis of N. viennensis EN76 floating filter-grown cells revealed upregulation of unique sets of genes for cell wall and extracellular polymeric substance biosynthesis, ammonia oxidation (including ammonia monooxygenase subunit C (amoC3) and multicopper oxidases), and defense against H2O2-induced oxidative stress. These genes may play a pivotal role in adapting AOA to air-exposed solid surfaces. Furthermore, the floating filter technique resulted in the enrichment of distinct soil AOA communities dominated by the “Ca. Nitrosocosmicus” clade. Overall, this study sheds light on distinct adaptive mechanisms governing AOA growth on air-exposed solid surfaces.
Introduction
Ammonia-oxidizing microorganisms (AOMs) play a crucial role in the global nitrogen cycle in soil environments [1, 2]. They include ammonia-oxidizing Archaea (AOA) and ammonia-oxidizing bacteria (AOB), which mediate the first and rate-limiting step of nitrification: ammonia (NH3) oxidation to nitrite (NO2−) [3, 4]. Another group of bacterial ammonia-oxidizers, known as complete ammonia-oxidizers (comammox), can oxidize NH3 to nitrate (NO3−) via NO2− in soil environments [5, 6].
Soil AOA constitutes a significant portion of the soil nitrification microbiome [1, 7]. AOA belongs to the class Nitrososphaeria, which is affiliated with the phylum Nitrososphaerota, formerly known as Thaumarchaeota [8–10]. They can be classified into four major orders: “Candidatus Nitrosocaldales”, Nitrosopumilales, “Ca. Nitrosotaleales,” and Nitrososphaerales [11]. Recently, Zheng et al. [12] reported “Ca. Nitrosomirales”, a novel order of the AOA widespread in terrestrial and marine environments. Nitrososphaerales is mainly a group of soil-dwelling AOA, also known as group I.1b [13]. They have some cultivated representatives and various subclusters that have not been cultured [14–16]. The isolated strains of Nitrososphaerales are obligate aerobes, mesophiles/moderately thermophiles with chemolithoautotrophic metabolism through ammonia oxidation and CO2 fixation [17–19].
Most soil microorganisms proliferate by colonizing mineral surfaces [20], within pores of soil aggregates. Surprisingly, the primary method for cultivating soil AOM remains the liquid culture system. Previous nitrification studies have mainly focused on strains isolated from liquid culture systems, and their physiology has been studied exclusively in liquid media. It is worth noting that AOM strains obtained through the liquid culture system may exhibit notable variations in their physiology when cultivated on a solid interface. These variations may arise from differences in substrate accessibility and chemical disparities at the solid–liquid interface. Hence, fundamental questions about their physiology within the pores of soil aggregates still need to be explored. To address this gap, we adopted the floating filter cultivation technique, previously described for enriching and isolating ubiquitous uncultivated bacteria [21–23], to simulate solid systems and investigate AOA adaptation to air-exposed solid surfaces. Although the floating filter cultivation technique has not been reported for AOA, a few studies have described the formation of microcolonies on membrane filters by AOB strains belonging to the genera Nitrosomonas and Nitrosospira [24, 25]. However, detailed information about their physiology when grown on these solid surfaces remains elusive.
Notably, the soil Nitrososphaerales typically grow in aggregates, implying the ability to extensively modify their cell to adhere to surfaces and form biofilm [13, 17, 19]. Thus, we compare the growth and physiological properties of soil-dwelling AOA from two phylogenetically distinct clades, Nitrosopumilales and Nitrososphaerales, along with a soil ammonia-oxidizing bacterium from the genus Nitrosomonas, on floating membrane filters and liquid media. We demonstrated that the physiology of AOA grown on air-exposed solid surfaces was different from those grown as freely suspended cells in liquid media. Using transcriptomic analysis, we revealed the distinct response mechanisms of these strains to the floating filter. Finally, the floating filter cultivation technique was demonstrated to enrich distinct soil surface-adapted AOA communities from agricultural soil.
Materials and methods
Cultivation in the liquid media
For this study, we selected representative strains of soil AOA of Nitrosopumilales and Nitrososphaerales clades, “Nitrosotenuis chungbukensis” MY2 [26] and Nitrososphaera viennensis EN76 [27], respectively, together with a soil ammonia-oxidizing bacterium, Nitrosomonas europaea ATCC 19718 [28, 29]. These pure AOM strains were cultivated in artificial freshwater medium (AFM) under optimal conditions, as previously described [26, 27, 30]. Briefly, pure cultures of N. viennensis EN76, “N. chungbukensis” MY2, and N. europaea ATCC 19718 were incubated in the dark without shaking at their optimum growth temperature and pH. The medium pH was 7.5 for the AOA strains and 7.8 for N. europaea ATCC 19718. Ammonium chloride (NH4Cl) (1 mM) from presterilized stocks was added as the sole energy source. The growth medium of N. viennensis EN76 and “N. chungbukensis” MY2 was always supplemented with sodium pyruvate (0.1 mM) as an H2O2 scavenger [27, 31]. The pH of the media was adjusted with sterile NaOH when necessary. The growth and ammonia oxidation activities of all strains were monitored by nitrite (NO2−) accumulation. NO2− concentration was measured colorimetrically using the Griess test, as previously described [26]. Cultures were regularly monitored for contamination of heterotrophic bacteria by checking turbidity in Luria-Bertani (LB) and Reasoner’s 2A (R2A) broth (1:10) [32].
Cultivation on the floating filters
We compared the growth and ammonia oxidation activities of three AOM strains, N. viennensis EN76, “N. chungbukensis” MY2, and N. europaea ATCC 19718 cells on floating filters and the control culture grown as freely suspended cells in AFM (hereafter referred to as control culture in liquid media). N. europaea ATCC 19718, N. viennensis EN76, and “N. chungbukensis” MY2, which had estimated cell densities of ~106 cells ml−1, 107 cells ml−1, and 108 cells mL−1, respectively, after oxidizing 1 mM ammonia, were serially diluted in 10-fold increments. A 1-ml aliquot of each dilution was used as an inoculum for floating filters and control culture in liquid media. Inoculum size hereafter is expressed as the number of cells in the 1-ml aliquot of each dilution. For the floating filter, prerinsed sterile nucleopore polycarbonate (PC) track-etched filters (Cytiva, Marlborough, MA, USA) with a diameter of 47 mm and a pore size of 0.1 μm were used for the experiment. The filter was mounted on the autoclaved filter holders (Nalgene Polysulfone Reusable Bottle Top Filter; Thermo Fisher Scientific, Waltham, MA, USA). Afterward, the 1-ml inoculum was mixed with 4 ml AFM and vacuum-filtered at 13 mbar. Then, the filter was aseptically placed on top of 10 ml AFM in a sterile polystyrene petri dish (60 × 15 mm) using sterile tweezers. The petri dishes were placed in a sealed container containing ambient air and incubated in the dark without shaking at 42°C, 30°C, and 25°C for N. viennensis EN76, “N. chungbukensis” MY2, and N. europaea ATCC 19718, respectively. A blank filter (uninoculated filter) was included as a negative control. The ammonia oxidation activities of all strains on floating filters were monitored by subsampling from the 10 ml AFM below the filter and measuring NO2− accumulation. Growth rates were calculated based on the assumption that NO2− production is correlated to the growth of ammonia oxidizers [33]. The specific growth rate (μmax) was calculated as the gradient of semilogarithmic plots of nitrite concentration versus time during exponential growth [34]. For biomass determination, the filter containing the grown cells was submerged into 10 ml AFM in the petri dish and scraped using a scraper (Cat. No. 90020). This procedure was repeated twice. The resulting cell suspension and the control culture grown in the liquid medium were harvested using centrifugation (13 000×g, 7 min, 25°C). The total cellular protein was extracted from the cell pellet using B-PER Protein Extraction Reagent (Thermo Fisher Scientific, Waltham, MA, USA) and quantified using the micro BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer’s instructions.
Microscopy
A fluorescence microscope was used to monitor the growth of AOM in microcolonies on the floating filter. The counterstain DAPI (4,6-diamidino-2-phenylindole) was added to reveal the size, morphology, and arrangement of the microcolonies of AOM on the filter. For staining, about 30 μl of 200 μg ml−1 DAPI was added on top of filters (cells side up), after which filters were incubated for 10 min and dried at 37°C in the dark. The dried filters were mounted on glass slides with coverslip and viewed at 1000-fold magnification. Microcolonies and cells were imaged using an Olympus BX61 fluorescence microscope with a U-MWU2 fluorescence mirror unit.
Transcriptomic analysis
Transcriptome analysis was performed on cells of N. viennensis EN76 (triplicates) and N. europaea ATCC 19718 (> triplicates) grown on floating filters and the control culture in liquid media. At first, we used a 10-ml AFM to cultivate cells of N. viennensis EN76 and N. europaea ATCC 19718 on floating filters (in petri dishes) and the control culture in liquid media (in cell culture flasks) under optimum conditions. Before the end of ammonia oxidation, once the NO2− had accumulated to ~0.7 mM, a further 1 mM NH4Cl was added to sustain exponential growth, and the cultures were scaled up to a vessel containing 300 ml AFM to obtain sufficient biomass. To scale up the volume of AFM for the floating filter-grown cells, the filters were transferred from petri dishes to bottles containing 300 ml fresh AFM with 1 mM NH4Cl. The total RNA was extracted from cells of the scaled-up filters, and the control cultures in liquid media after ~0.6 mM ammonia was oxidized. For the control culture in liquid media, cells were harvested by filtration, using mixed cellulose esters (MCEs) filters (Advantec Mfs Inc, Dublin, CA, USA) with a diameter of 47 mm and pore size of 0.2 μm. The filters were individually processed by grinding with a mortar and pestle using liquid nitrogen. Total nucleic acid (DNA and RNA) was extracted for each sample using a modified CTAB reagent [35]. The extracted nucleic acids were passed through the AllPrep DNA/RNA Mini Kit (Qiagen, Germany) to separate RNA from DNA, following the manufacturer’s instructions. During the RNA purification process, the DNA were digested on-column using the DNase I digestion set (Thermo Fisher Scientific, Waltham, MA, USA). The absence of residual genomic DNA was verified via PCR amplification using universal 16S rRNA gene sequences for 30 cycles.
RNA quality check was performed with the Agilent 2100 Expert Bioanalyzer (Agilent), and cDNA libraries were prepared directly from total RNA without rRNA removal using the Nugen Universal Prokaryotic RNA-Seq Library Preparation Kit. The cDNA libraries were sequenced via NovaSeq6000 (Illumina) at LabGenomics (Seongnam, Korea). Reads quality was assessed by FastQC (v0.11.8) [36]. For trimming reads, Trimmomatic (v0.36) [37] was used with the options: SLIDINGWINDOW:4:15 LEADING:3 TRAILING:3 MINLEN:38 HEADCROP:13. Reads mapped to N. viennensis EN76 and N. europaea ATCC 19718 rRNA sequences were removed with SortMeRNA (v2.1) [38]. Next, the remaining reads were aligned to the genome of each strain using Bowtie2 (v2.4.4) [39], and the reads mapped to each gene were counted using HTSeq (v0.12.3) [40]. Expression values are presented as transcripts per kilobase million (TPM). The statistical analysis of differentially expressed genes in cells grown on floating filters and liquid media was performed using the DESeq2 Bioconductor package (version 1.42.1) [41] in the R software environment (version 4.3.1) [42].
ROS scavenger experiment
We investigated the growth response of N. viennensis EN76 cells on floating filters in the presence of different ROS scavengers. Catalase (10 U mL−1) and varying concentrations of pyruvate (0, 0.1, and 1 mM) were used against H2O2-induced oxidative stress on the floating filter-grown cells compared to the control culture in liquid media supplemented with 0.1 mM pyruvate. Cells of N. viennensis EN76 with inoculum sizes of ~107 and 106 cells were used for the experiments.
Cultivation of AOA communities from agricultural soil
A composite bulk soil sample was collected from an experimental agricultural station in Chungbuk National University Republic of Korea (127°27′18.5″E, 36°37′29.8″N) during the fallow period in March 2023. The soil had not been exposed to fertilization at that time. Six individual 30 cm subsoils were collected within a 10 × 10 m2 area from plots at the agricultural station. The subsoils, which were loamy sand, were combined and transported to the laboratory to be stored at 4°C before being used for inoculation. The properties of the soil were as follows: loam texture (sand, 51%; silt, 33%; and clay, 16%); water content, 4.2%; pH, 6.0; total organic carbon, ˂0.1 g kg−1; total nitrogen, ˂0.01%; total ammonia, 6.0 mg kg−1; total phosphate, 174.8 mg kg−1; and cation exchange capacity, 12.1 cmol kg−1.
For enrichment cultures, AFM was prepared with 4 g of CaCO3 per litre (L) medium. One gram (1 g) of soil was added into 9 ml AFM (supplemented with the CaCO3 particles), vortexed briefly, and serially diluted in 10-fold increments (i.e. 10−1 to 10−4). For comparison, 1 ml of the 10−2-fold diluted soil suspension was used as inoculum for floating filters and control culture in liquid media. The final pH of the medium was 7.5. The protocol used to cultivate pure AOM strains on floating filters (described earlier) was adopted for enriching soil AOM on floating filters. The petri dishes were placed in a sealed container containing ambient air and incubated in the dark at 30°C without shaking. Nitrification activity was monitored by measuring NO2− and NO3− accumulation. NO2− and NO3− concentrations were quantified colorimetrically using the Griess and VCl3/Griess reagents, respectively [43]. For successive transfer of the enrichment cultures, the filter containing the grown cells was submerged into 10 ml AFM in the petri dish and scraped using a scraper (as mentioned earlier). This procedure was repeated twice and 10% of the resulting cell suspension was used as inoculum for new filters. For comparison, ca. 10% of the control culture in liquid media was also repeatedly transferred for successive enrichment. Following four rounds of transfers, the microbial communities enriched on floating filters and liquid media were examined using 16S rRNA gene amplicon sequencing.
16S rRNA gene amplicon sequencing
For analysis of microbial communities enriched from the agricultural soil, DNA extracted from the enrichment cultures and original inocula were used as templates for 16S rRNA gene amplicon sequencing analysis. A modified CTAB method [35] was employed in extracting high molecular weight genomic DNA from the enrichment cultures grown on floating filters and liquid media. In brief, biomass obtained from the cultures was treated with CTAB and sodium dodecyl sulfate (SDS) extraction buffer, incubated for 30 min at 65°C with occasional mixing, and centrifuged at 8000×g, 10 min, 25°C. The supernatant obtained was repeatedly purified with an equal volume of chloroform/isoamyl alcohol (24:1). The extracted DNA was precipitated with 0.6 volume of 2-propanol, and the pelleted DNA was washed twice with 70% ethanol, allowed to air dry, and resuspended in TE buffer (10 mM Tris, pH 8, 1 mM EDTA). The extracted DNA concentrations were measured with a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and the quality was assessed on a 1% (w/v) agarose gel.
The hypervariable V4-V5 region of the 16S rRNA gene was amplified with the primer pair 515F/926R [44] and sample indexing adapters (Nextera XT index kit). PCR amplifications were conducted via the following steps: 3 min heating step at 95°C, followed by 25 cycles at 95°C for 45 s, 50°C for 45 s, 72°C for 90 s, and 72°C for 5 min. The PCR product was purified using the Labopass purification kit (Cosmo Genetech, South Korea), and the quantity obtained was measured with a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The quality of the PCR product was assessed on a 1.5% (w/v) agarose gel. The library was sequenced using the Illumina MiSeq (2 × 300 bp) platform at Macrogen (Seoul, Republic of Korea). The raw sequence reads were analyzed using the QIIME (QIIME2-2023.2) [45] pipeline implemented with tools for quality control (Cutadapt) [46], denoising and pair read merging using DADA2 (version 1.6.0) [47], and for amplicon sequence variants (ASVs) inference. Taxonomic ranks were assigned to the inferred ASVs using the SILVA ribosomal reference database (SILVA release 132) [48].
Phylogenetic analysis
For phylogenetic analysis of the 16S rRNA gene sequence, we retrieved representative nucleotide sequences of related taxa from the National Center for Biotechnology Information (NCBI) database. In addition, for the phylogenetic analysis of the sigma-70 protein, we obtained the amino acid sequences from the genome of N. europaea ATCC 19718, as well as representative amino acid sequences from four diverse bacteria: Escherichia coli, Caulobacter vibrioides, Bacillus subtilis, and Pseudomonas aeruginosa, all sourced from their genome sequences on NCBI. The 16S rRNA gene sequences and the amino acid sequences of the sigma-70 protein were aligned separately with MAFFT (v7.313) with the L-INS-I method [49]. Maximum-likelihood trees were inferred with IQ-TREE (v2.0.6) [50] and the best-fitting model was selected using the fast model selection (m MFP) [51]. The constructed trees were visualized using iTOL v6 [52].
Statistical analysis
All statistical analyses were conducted using the R statistical software (v4.3.1) and R Studio (v2022.02.3). A one-way or two-way analysis of variance (ANOVA) with Tukey’s honest significant differences post hoc multiple comparisons test (p < 0.05) was used to distinguish statistical significance. Microbial diversity analysis and data visualization were performed using the R packages phyloseq (v1.26.0) [53], vegan (version 2.6–8) [54], and ggplot2 (v3.1.0) [55]. Processed amplicon sequence reads were imported using phyloseq [53]. Nonmetric multidimensional scaling (NMDS) analysis, based on Bray–Curtis dissimilarity metrics, was used in vegan (version 2.6–8) [54] to compare the microbial communities between the samples. The ordination analysis patterns were statistically tested using permutational multivariate analysis of variance (PERMANOVA) and analysis of similarity (ANOSIM) with adonis2 and vegdist, respectively, which are part of the vegan (version 2.6–8) packages in R [54]. Indicator species were identified using the indval function of the labdsv package (v2.0–1) [56]. Samples from the floating filter culture and liquid culture were treated as a group for indicator species analysis.
Results and discussion
Growth on floating filters
All the strains with varying cell densities were cultivated on floating filters, with the control culture in liquid media. The ammonia oxidation activity of N. viennensis EN76 cells with an inoculum size of ~107 cells decreased on floating filters compared with the control culture in liquid media (Fig. 1A and B). Furthermore, the ammonia oxidation activity of N. viennensis EN76 with an inoculum size of 105 cells was significantly repressed on floating filters with less than < ca. 0.05 mM ammonia oxidized in 20 days (Fig. 1A). However, inoculum sizes had no significant effect on the ammonia oxidation activity and the specific growth rate (μmax) (ranging from 0.65 ± 0.03 d−1 to 0.68 ± 0.01 d−1) of the control culture in liquid media, as shown in the subfigure at the upper right panel. Nonetheless, the time required for complete oxidation of 1 mM ammonia almost proportionally increased with inoculum size (Fig. 1B). Interestingly, differences in inoculum size only affected the ammonia oxidation activity of N. viennensis EN76 cells grown on floating filters. On the contrary, “N. chungbukensis” MY2 cells showed no ammonia oxidation activity on floating filters compared to the control culture in liquid media (Fig. 1C and D). The ammonia oxidation activity of N. europaea ATCC 19718 cells grown on floating filters (with μmax ranging from 0.71 ± 0.01 d−1 to 0.75 ± 0.00 d−1) was not significantly different from the control culture in liquid media (with μmax ranging from 0.75 ± 0.02 d−1 to 0.76 ± 0.02 d−1). Moreover, reducing the inoculum sizes by 10-fold increments (i.e.,~105 and 104 cells) did not repress the ammonia oxidation activity of N. europaea ATCC 19718 cells grown on floating filters, as the μmax of different inoculum sizes (refer to the upper right panel) was comparable (Fig. 1E and F). A detailed description of the statistical significance is provided in Supplementary Table S1.
Comparison of ammonia oxidation activities of AOM grown on floating filters and the control culture in liquid media. The line graphs (A, C, E) and (B, D, F) represent floating filters and liquid cultures, respectively. N. viennensis EN76 (A and B), “N. chungbukensis” MY2 (C and D), and N. europaea ATCC 19718 (E and F) grown on floating filters and liquid media, respectively, with varying inoculum size. The subfigure in the upper right panel shows the specific growth rate (μmax) with different inoculum sizes. The symbols * and × indicate cultures for which growth rates cannot be calculated due to the absence of a log phase and no growth, respectively. All experiments were performed in triplicates. Data are presented as mean ± standard deviation (SD) (n = 3), and the error bars are hidden when they are smaller than the width of the symbols. To avoid overlapping symbols, the value was shifted by −0.03 and 0.02 in (A) for the experiment with 106 cells and 105 cells, respectively, and by −0.09, 0.06, or 0.02 in (C) for the 108 cells, 107 cells, and 106 cells, respectively.
Considering no significant decrease in ammonia oxidization activity of N. europaea ATCC 19718 cells on floating filters, it can be inferred that any potential limitation in nutrient transport through the filters, which have ~108 pores/cm2, is likely insignificant, at least for N. europaea ATCC 19718. Since the number of pores on the filters is greater than the size of inoculated cells (<108 cells/filter; filtered area approx. 11.3 cm2), it is improbable that the pores will become clogged by the inoculated cells. The effect of filter materials on the ammonia oxidation activity of N. viennensis EN76 cells with an inoculum size of ~107 cells was tested on other filters: polyvinylidene fluoride (PVDF), polytetrafluoroethylene (PTFE), and MCEs. We observed comparable ammonia oxidation activities between the alternative filters (as mentioned above) and polycarbonate (PC) filters, indicating no significant inhibitory effect of filter materials (Fig. S1). Thus, the reason for the severe growth retardation in the lower inoculum sizes (~106 and 105 cells) of N. viennensis EN76 cells on the surface of the filter is unclear. It is conceivable that an unknown cell density-dependent cooperative interaction and cellular responses [57] might be responsible. Additionally, the volatilization of essential metabolic intermediates of ammonia oxidation, such as the gaseous NOx (HONO + NO + NO2) [58, 59], may also have a serious impact on lower-density cultures on floating filters, which requires further investigation. Furthermore, we investigated whether the reduced growth of N. viennensis EN76 cells on floating filters (~107 cells) was partly due to substances leaching out of the polystyrene petri dishes. The result indicated that the ammonia oxidation activity of N. viennensis EN76 cells in the liquid media, whether in the cell culture flasks or petri dishes, is comparable and notably different from their activity on floating filters (Fig. S2).
Using fluorescence microscopy, we observed microcolonies formation by N. viennensis EN76 with higher inoculum size (~107 cells) on the filters before and after oxidation of 1 mM ammonia (Fig. 2A and B), respectively. The size of microcolonies was further increased after oxidizing an additional 1 mM ammonia supplied into the AFM beneath the filters (Fig. 2C). However, the microcolony formation of N. viennensis EN76 with lower inoculum size (~106 cells) was poor compared with higher inoculum size (~107 cells) during the same incubation period (ca. 20 days) (Fig. 2D), corresponding with the observed decreased ammonia oxidation activity (refer to Fig. 1A). Furthermore, cellular protein quantification revealed that N. viennensis EN76 cells grown on floating filters produced less biomass compared to the control culture in liquid media despite oxidizing the same 1 mM ammonia (Fig. S3), implicating stress conditions for AOA grown on the filter.
Fluorescent micrographs of N. viennensis EN76 microcolonies grown on floating filters. Micrographs of N. viennensis EN76 with an inoculum size of ~107 cells on polycarbonate filter (A) before and (B) after 1 mM ammonia oxidation (refer to Fig. 1A) and (C) microcolonies of N. viennensis EN76 after oxidation of an additional 1 mM ammonia. (D) Microcolonies of N. viennensis EN76 with an inoculum size of ~106 cells after 20 days of incubation (refer to Fig. 1A). The cells were stained with DAPI for 10 min and dried at 37°C on a glass slide.
Effect of CaCO3 and filtration inoculation
Calcium carbonate (CaCO3) particles are often included in AOM cultures (liquid media) as a buffer against acid stress [60] and may increase the available surface that induces biofilm or microcolony formation in AOM [61, 62]. Owing to their deliquescent properties, CaCO3 particles surrounding the cells on the filters may act as a protective shield against dryness. Interestingly, we observed that inoculating lower inoculum size (~106 cells) of N. viennensis EN76 with CaCO3 particles on the filters improved the ammonia oxidation activity compared to cells on filters without CaCO3 particles (Fig. 3A). In contrast, the growth of lower inoculum size (~107 cells) of “N. chungbukensis” MY2 on floating filters could not be recovered even with CaCO3 particles supplementation (Fig. 3B), indicating their inability to grow on the air-exposed filters or permanent damage of cells caused by the vacuum filtration process.
Ammonia oxidation activities of N. viennensis EN76 and “N. chungbukensis” MY2 cells under different growth conditions. N. viennensis EN76 (A) and “N. chungbukensis” MY2 (B) with inoculum sizes of ~106 cells and ~ 107 cells, respectively, were used for the experiments. The subfigure in the upper right panel shows the specific growth rate (μmax) under different culture conditions: 1 = filter without CaCO3; 2 = filter with CaCO3; 3 = liquid media without CaCO3; 4 = inverted filter without CaCO3. The symbols * and × indicate cultures for which growth rates cannot be calculated due to the absence of a log phase and no growth, respectively. All experiments were performed in triplicates. Data are presented as mean ± SD (n = 3), and the error bars are hidden when they are smaller than the width of the symbols. To avoid overlapping symbols, the value was shifted by −0.02 in (A) for the filter without CaCO3 and by −0.09, 0.05, or 0.01 in (B) for the filter without CaCO3, filter with CaCO3, and inverted filter culture, respectively.
To assess the viability of “N. chungbukensis” MY2 cells after filtration, filters on which the cells were deposited were inverted, exposing the cells directly to the AFM beneath the filter (inverted floating filter). The “N. chungbukensis” MY2 cells (~107 cells) showed no ammonia oxidation activity on inverted floating filters (Fig. 3B). However, the μmax of N. viennensis EN76 cells (~106 cells) on inverted floating filters (0.59 ± 0.00 d−1) was comparable to the control culture in liquid media (0.66 ± 0.02 d−1) (refer to Fig. 3A). Nevertheless, there was a slight decrease in the μmax of N. viennensis EN76 cells on inverted floating filters (as mentioned earlier) and filters with CaCO3 particles (0.44 ± 0.01 d−1). The significant decrease in the ammonia oxidation activity and μmax of N. viennensis EN76 cells on inverted floating filters and air-exposed floating filters without CaCO3 particles (refer to Fig. 3A and Supplementary Table S2) further signifies a unique life strategy for AOA cells growing on air-exposed solid surfaces. However, it is conceivable that the cells on the air-exposed floating filter are in a thin liquid layer, which aids in transporting nutrients from the AFM beneath the filters to the microcolonies. Together, our findings indicate that cells of “N. chungbukensis” MY2 of the Nitrosopumilales lost viability and are more vulnerable to physical stress damage than cells of N. viennensis EN76 of the Nitrososphaerales. This notion is supported by results from previous studies wherein “N. chungbukensis” MY2 and Nitrosopumilus maritimus SCM1 of Nitrosopumilales lost their ammonia oxidation activities after the cells were concentrated by filtration or centrifugation [26, 63]. Accordingly, we observed ammonia oxidation activity from “N. chungbukensis” MY2 cells on floating filters when inoculated on filters using ambient gravitational force without vacuum suction (Fig. S4). Nonetheless, the observed ammonia oxidation activity of these cells was lower than the control culture in liquid media. This indicates that cells of “N. chungbukensis” MY2 may not be well adapted to solid surfaces, as it has been reported that AOA of Nitrosopumilales have limited capability to make extracellular polymeric substances (EPS), essential for biofilm or microcolony formation compared to AOA of Nitrososphaerales [64].
Transcriptomic analysis of N. viennensis EN76
Considering the differences in ammonia oxidation activities (see above) of AOA cells grown on floating filters and the control culture in liquid media, we analyzed and compared the transcriptome of N. viennensis EN76 cells grown on floating filters and liquid media to better understand how AOA adapt to the air-exposed solid surfaces. Out of the 2944 genes identified in the transcriptomes of N. viennensis EN76 (Supplementary Table S3), 741 (25.17% of detected genes) exhibited significant differential expression with a log2FC > 1 and FDR < 0.05 (Supplementary Table S4). Among the differentially expressed genes, 475 (64.1%) were significantly upregulated, and 266 (35.9%) were significantly downregulated in N. viennensis EN76 floating filter-grown cells. This differential gene expression indicates notable physiological differences between AOA cells grown as microcolonies on floating filters and the control culture in liquid media. Most differentially expressed genes in cells grown on floating filters could be functionally categorized in (i) cell wall and extracellular polymeric substance (EPS) biosynthesis, (ii) inorganic ion transport and metabolism, (iii) posttranslational modification, protein turnover, and chaperones, (iv) carbohydrate transport and metabolism, and (v) signal transduction mechanisms. The “function unknown” category contains the largest fraction of upregulated genes in floating filter-grown cells (Supplementary Table S4 and Fig. S5). Differentially expressed genes critically involved in the adaptation to growth on the filters were further analyzed.
Cell wall and EPS biosynthesis
The genes involved in EPS biosynthesis, e.g., genes encoding glycosyl transferase, exported polysaccharide deacetylase, sialidase-neuraminidase family protein, methyltransferase, N-acetyltransferase, mannosyltransferase, sulfotransferase, and xylanase/chitin deacetylase [64], which often reside in clusters, were upregulated in floating filter-grown cells of N. viennensis EN76 (Supplementary Table S5). Some of these enzymes belong to the family of transporters (TC#4.D.1 and 4.D.2) that couple polysaccharides biosynthesis with translocation across the membrane. EPS production is essential for microcolony and biofilm development [65]. Thus, the functions of these enzymes may be critical for cell surface modification required for solid surface adaptation, thereby promoting survival by establishing interaction between cells and protecting cells from air-exposed solid surfaces. Besides EPS biosynthesis, genes encoding cell surface-associated proteins, such as hemolysin, pilin, and surface anchor family protein, were downregulated in floating filter-grown cells (Supplementary Table S5).
Ammonia oxidation
Most of the genes of different ammonia monooxygenase subunits: amoA, amoB, amoX, amoY, amoC1, amoC2, amoC4, amoC5, and amoC6 were constitutively expressed (Supplementary Table S6). Interestingly, the transcript of amoC3 was highly upregulated (> 100-fold) in floating filter-grown cells of N. viennensis EN76 (Supplementary Table S6). It is important to note that amoC3 and amoC6 in N. viennensis EN76 are the core amoC COG in Nitrososphaerales [64]. Hodgskiss et al. [66] reported that N. viennensis EN76, like most other soil-dwelling AOA, encodes multiple homologs of the amoC gene and pinpointed amoC6 as the primary homolog within the AMO enzyme complex. Depending on the environmental conditions, the various amoC homologs in the genome might provide different activity profiles to the AMO complex. Previous transcriptional studies showed that additional monocistronic amoC subunits found in some terrestrial AOB aid in maintaining AMO enzyme stability during stressful conditions [67–69]. Together, our findings suggest a pivotal role of the amoC3 gene in the ammonia oxidation activity of N. viennensis EN76 cells grown on air-exposed solid surfaces.
The genome of N. viennensis EN76 encodes four of the type C two-domain multicopper oxidases (2dMCOs), with their copper-binding site (T1 center) overlaid onto the corresponding region in other type C 2dMCOs (Fig. S6). Notably, transcripts of MCOs were differentially expressed in N. viennensis EN76 cells grown on floating filters. MCO (NVIE_RS08635) and MCO (NVIE_RS12885) were significantly upregulated (>5- and 29-fold, respectively), while MCO (NVIE_RS00300) was downregulated in floating filter-grown cells. On the contrary, MCO (NVIE_RS09330) showed constitutive expression (Supplementary Table S6). In AOA, MCOs are predicted to participate in the archaeal HURM (hydroxylamine: ubiquinone redox) module, catalyzing the oxidation of hydroxylamine (the first product of ammonia oxidation) [70, 71]. Thus, it is conceivable that the MCOs (NVIE_RS08635 and NVIE_RS12885) might be involved in hydroxylamine oxidation in N. viennensis EN76 cells grown on air-exposed floating filters. N. viennensis EN76 cells grown on the filters seem to modify AMO and HURM systems to generate energy for surviving on solid surfaces.
Energy conservation
It has been suggested that electrons released into the electron transport chain (ETC) by hydroxylamine oxidation in AOA are transferred to the quinone/quinol pool (Q/QH2) via plastocyanin-like electron carriers [71, 72]. The floating filter-grown cells of N. viennensis EN76 showed significant upregulation (> 2-fold) of genes encoding the plastocyanin-like electron carriers and ferredoxins involved in electron transport (Supplementary Table S7). In addition, the genes encoding the PetC (NVIE_RS09625) and CoxA2 (NVIE_RS00660) of Complex III and IV, respectively, were upregulated (> 4- and 15-fold, respectively) (Supplementary Table S7). Electrons generated from ammonia oxidation are transferred to the bimetallic cytochrome a3/CuB active site in coxA2, where the reduction of dioxygen molecules to water occurs [73–75]. Furthermore, genes encoding the AtpC (NVIE_RS11020) and AtpD (NVIE_RS11000) of Complex V (ATP synthase), which are integral components of the catalytic site that utilizes proton motive force (PMF) for ATP synthesis, were downregulated in N. viennensis EN76 floating filter-grown cells (Supplementary Table S7). Meanwhile, other subunits of the ATP synthase showed constitutive expression. In the ETC, complex I (NADH-quinone oxidoreductase) is used for the reverse transport of electrons to reduce NAD+ in AOM [76–78]. Genes encoding six subunits (NuoN; NVIE_RS05590, NuoL; NVIE_RS05595, NuoM; NVIE_RS05600, NuoJ; NVIE_RS05610, NuoI; NVIE_RS05615, NuoA; NVIE_RS05640) of complex I were downregulated in N. viennensis EN76 floating filter-grown cells (Supplementary Table S7). The downregulation of these genes suggests reduced NADH production for anabolism. Thus, it is plausible that N. viennensis EN76 uses the activity of ETC to maintain PMF and sustain membrane potential for stress responses in cells grown on floating filters. The increased use of PMF has been reported in some bacteria under several starvation/stress conditions [79–81]. Differential expressions of genes involved in ammonium transport (Supplementary Table S6) and cell division (Supplementary Table S8) are described in Supplementary Note 1.
Oxidative stress responses/inorganic nutrient homeostasis
Maintaining iron homeostasis is crucial in responding to oxidative stress caused by the Fenton reaction, where Fe(II) reacts with H2O2, generating a highly reactive and toxic hydroxyl radical (OH·) [82]. This reaction is primarily regulated by the miniferritins family of proteins known as Dps (DNA-binding protein from starved cells) [83, 84]. Dps proteins bind and store Fe(II), ferroxidizing it effectively using H2O2 as the oxidant while simultaneously detoxifying H2O2 [83, 85], thus preventing OH· production. Two of the genes encoding Dps proteins in N. viennensis EN76 (Dps3; NVIE_RS09700 and Dps4; NVIE_RS13730) were upregulated (> 6-fold) in floating filter-grown cells (Supplementary Table S9). Dps3 (NVIE_RS09700) in N. viennensis EN76 shares 68.16% identity with the Dps protein (WP_009989805.1) in Saccharolobus solfataricus (formerly referred to as Sulfolobus solfataricus). On the other hand, N. viennensis EN76 Dps4 (NVIE_RS13730) shares 24% and 40.76% identities with Dps protein from E. coli (WP_000100800.1) and P. aeruginosa PAO1 (NP_249653.1), respectively. Maaty et al. [86] reported high upregulation of the genes encoding Dps protein in S. solfataricus in response to H2O2-induced oxidative stress. Notably, the genes encoding Dps protein in E. coli and P. aeruginosa PAO1 are known to protect cells from H2O2-mediated oxidative stress [85, 87]. Taken together, we suggest that the increased expression of the genes encoding Dps proteins in N. viennensis EN76 floating filter-grown cells may play a role in iron homeostasis by reducing the free iron that could catalyze the Fenton reaction.
Effect of H2O2 scavengers on ammonia oxidation activity of N. viennensis EN76 cells grown on floating filters. Growth of N. viennensis EN76 cells with an inoculum size of (A) ~107 cells and (B) ~106 cells on floating filters provided with different concentrations of pyruvate and catalase (10 U ml−1), as compared to the standard pyruvate concentration used in liquid media. The subfigure in the upper right panel shows the specific growth rate (μmax) under different culture conditions: 1 = 0 mM pyruvate; 2 = 0.1 mM pyruvate; 3 = 1 mM pyruvate; 4 = catalase (10 U ml−1); 5 = 0.1 mM pyruvate (liquid media). The symbols * indicate cultures for which growth rates cannot be calculated due to the absence of a log phase. All experiments were performed in triplicates. Data are presented as mean ± SD (n = 3), and the error bars are hidden when they are smaller than the width of the symbols.
The gene encoding a rubrerythrin-like protein in N. viennensis EN76 (NVIE_RS11805) was upregulated (> 2-fold) in floating filter-grown cells (Supplementary Table S9). Lumppio et al. [88] reported that the rubrerythrin protein in a sulfate-reducing anaerobic bacterium, Nitratidesulfovibrio vulgaris (formerly referred to as Desulfovibrio vulgaris), protects against exposure to air and H2O2-induced oxidative stress. Rubrerythrin protein can function as the terminal component of an NADH peroxidase, catalyzing the reduction of H2O2 to water [88]. The gene encoding the rubrerythrin-like protein in N. viennensis EN76 (NVIE_ RS11805) shares 38.71% and 30.65% identity with two rubrerythrin proteins (WP_010937330.1 and WP_010940353.1), respectively, in N. vulgaris. Previous studies [86, 89, 90] also showed that rubrerythrin proteins were upregulated in Methanothermobacter thermautotrophicus, Porphyromonas gingivalis, and S. solfataricus in response to H2O2-induced oxidative stress. In addition, genes encoding thioredoxin (TrxA1; NVIE_RS13995 and TrxA2; NVIE_RS14355), thioredoxin-like domain/NHL repeat-containing protein (ResA; NVIE_RS00045), and peroxiredoxin-like protein (NVIE_RS05335) in N. viennensis EN76 were upregulated (> 2-fold) in floating filter-grown cells (Supplementary Table S9). Collectively, it is tempting to infer that the upregulation of genes encoding rubrerythrin-like, thioredoxin, and peroxiredoxin-like proteins in floating filter-grown cells of N. viennensis EN76 might be playing an important role in protecting the cells against H2O2-induced oxidative stress on floating filters. In contrast, the genes encoding alkyl hydroperoxide reductase (NVIE_RS05690 and NVIE_RS06670) and superoxide dismutase (NVIE_RS14475) in N. viennensis EN76 showed constitutive expression.
Copper is a vital trace element for aerobic organisms [91]. Bacteria possess a periplasmic CopC protein that binds copper and delivers it to the inner membrane CopD protein, which potentially transports copper into the cytoplasm [91, 92]. The two Cop proteins identified in the genome of N. viennensis EN76 are fusions of CopC and CopD domains, and their copper-binding site is highly conserved when aligned with other CopC and CopD representative sequences (Fig. S7). The genes (NVIE_RS06945 and NVIE_RS06955) encoding these proteins were upregulated (> 4-fold) in N. viennensis EN76 cells grown on floating filter grown (Supplementary Table S9). Intriguingly, E. coli cells exposed to high levels of copper showed less sensitivity to H2O2-induced DNA damage [93]. Recently, Guerra et al. [94] reported that cupric (Cu2+) ions occupy specific binding sites in Dps, thus exerting a significant rate-enhancing effect on ferroxidation reaction in Marinobacter nauticus. This suggests that importing copper could serve as an effective mechanism for preventing or minimizing damage triggered by iron-induced OH· via classical Fenton reactions (as mentioned earlier). Similarly, a previous proteome study of three Nitrosopumilus strains (Nitrosopumilus maritimus, N. adriaticus, N. piranensis) revealed a relatively high abundance of Cop proteins in cells exposed to H2O2, implying a novel strategy for preventing oxidative damage through copper accumulation [95]. Together, it is tempting to propose that the upregulation of the copC/copD genes in N. viennensis EN76 floating filter-grown cells might represent a strategy for copper acquisition and a unique approach to prevent oxidative cell damage on the air-exposed floating filters.
The genes encoding subunits A and B (KdpA; NVIE_RS12995 and KdpB; NVIE_RS12990) of the potassium translocating ATPase (Kdp) were upregulated (> 3- and 2-fold, respectively) in N. viennensis EN76 floating filter-grown cells, while the gene encoding subunit C (KdpC; NVIE_RS12985) was constitutively expressed (Supplementary Table S9). The Kdp complex is a hybrid system for K+ transport that combines both the potassium (K+) superfamily transporters (KdpA) and the P-type ATPases (KdpB) [96, 97]. During the assembly of the Kdp complex, KdpC binds to KdpA to stabilize the complex, then K+ is transported from the KdpA subunit to the binding site in the KdpB subunit, where it is released to the cytosol [97]. In cells, the difference in K+ concentration across the plasma membrane plays a significant role in establishing the membrane potential, which is essential for regulating intracellular pH and generating the turgor pressure required for cell growth and division [98, 99]. In acidophiles the high affinity K+ uptake system, Kdp complex, is especially known to play critical roles in acid stress responses [100]. Nitrification causes acidification of media, which might be intense on floating filters. As observed above (refer to Fig. 3A), CaCO3 particles increased the ammonia oxidation activity of N. viennensis EN76 cells on floating filters. Thus, K+ homeostasis by the Kdp system is conceivably important for the acid adaptation of AOA on solid surfaces. It is notable that the representative acidophilic soil AOA, Nitrosotalea clade, as well as members of acid-adapted non-AOA soil Nitrososphaerota, Gagatemarchaeaceae, harbor the Kdp system [101, 102].
H2O2 scavenger effect
The upregulation of genes involved in responding to H2O2-induced oxidative stress in N. viennensis EN76 cells grown on floating filters prompted us to investigate the effect of H2O2 scavengers on the floating filter culture. Typically, a concentration of 0.1 mM pyruvate is usually sufficient for scavenging H2O2 in liquid cultures of catalase-negative soil and marine AOA [18, 31]. To enhance H2O2 scavenging activity in the floating filter culture, we provided a higher concentration of pyruvate (1 mM) and catalase (10 U mL−1) into the media beneath the filter. Our results revealed that a higher concentration of pyruvate (1 mM pyruvate) and 10 U ml−1 catalase significantly increased the ammonia oxidation activity of N. viennensis EN76 cells grown on floating filters (Fig. 4A and B), indicating an improved H2O2 scavenging activity. Noticeably, the ammonia oxidation activity and μmax of N. viennensis EN76 cells with an inoculum size of ~106 cells increased significantly due to higher H2O2 scavenging activity (Fig. 4B) A detailed description of the statistical significance is provided in Supplementary Table S10. Similarly, Bayer et al. [95] observed that higher initial cell density in Nitrosopumilus cultures could overcome H2O2-induced growth arrest in liquid media compared to cultures with lower cell density. This observation suggests that AOA cells exhibit a density-dependent cooperative defense against H2O2-induced oxidative stress. Together, the severe repression of ammonia oxidation activity of N. viennensis EN76 cells on floating filters (refer to Fig. 1A) may be attributed, in part, to H2O2-induced stress. Our findings suggest that the increased production of H2O2 and/or enhanced Fenton-like reactions on the solid surface may contribute to the niche differentiation of soil AOA.
Transcriptomic analysis of N. Europaea ATCC 19718
We also compared the transcriptome of N. europaea ATCC 19718 cells grown on floating filters and the control culture in liquid media (Supplementary Table S11). Among the 2619 genes identified in the transcriptome of N. europaea ATCC 19718, only 70 genes exhibited significant differential expression with a log2FC > 1 and FDR < 0.05. Prominently, genes encoding the sigma-70 (σ 70) family of the RNA polymerase (i.e., fecI), fecR-like genes, and the fecA (TBDT: TonB-dependent receptor) were the most abundant in upregulated genes in N. europaea ATCC 19718 floating filter-grown cells (Supplementary Table S12). Some genes involved in peptidoglycan synthesis and cell division were downregulated. The differential expression of a relatively small fraction of genes in N. europaea ATCC 19718 suggests that differences in the physiology of cells grown on floating filters and the control culture in liquid media are less significant when compared with N. viennensis EN76. This is consistent with the lack of notable differences in the ammonia oxidation activity of N. europaea ATCC 19718 between cells grown on floating filters and the control culture in liquid media (refer to Fig. 1E and F). Further description of the differentially expressed genes and the phylogenetic analysis of the nine upregulated σ 70 protein sequences in N. europaea ATCC 19718 cells grown on floating filters is provided and described in Supplementary Note 2 and Fig. S8, respectively.
Cultivation of soil AOA on floating filter
The growth of AOA on floating filters provided valuable insights into their growth within the pores of soil aggregates. Thus, we adopted this technique to grow surface-adapted AOA from agricultural soil and compared it with those grown in liquid media. To focus on AOA communities, allylthiourea (ATU, 50 μM), an inhibitor of bacterial ammonia-oxidizers [103, 104], was added to the growth media. Following the oxidation of about 0.7 mM ammonia in the first culture on floating filters and liquid media, 10% of the biomass was successively transferred to fresh filters and liquid media, respectively (Fig. 5A and B). Notably, there was no significant accumulation of NO2− as an intermediate; instead, NO3− was the product detected.
Cultivation of AOA from agricultural soil using the floating filter and liquid media cultivation technique. Experiments were conducted with the addition of 50 μM of ATU to inhibit the growth of AOB and comammox. Accumulation of NO2− + NO3− indicates ammonia oxidation activity in the cultures. The line graphs (A and B) represent the ammonia oxidation activities of two biological replicates of floating filters and liquid cultures, respectively, during four successive culture transfers. A NMDS analysis of the overall nitrifiers’ ASVs enriched on floating filters and liquid media is shown in (C). The letters and numbers indicate the following: L = liquid culture, F = floating filter culture, a & b = two biological replicates, 1 = first culture, 2 = 2nd culture, 3 = 3rd culture, 4 = 4th culture, and 5 = 5th culture.
After four rounds of transfers, 16S rRNA gene amplicon sequencing was used to examine the nitrifier communities enriched on floating filters and liquid media. A NMDS plot using the 16S rRNA gene ASVs revealed significant variation (ANOSIM test, stress = 0.1540, R = 0.8062, P = .001) in nitrifier communities enriched on floating filters and liquid media during the transfers (Fig. 5C). The relative abundance of nitrifiers’ ASVs on floating filters was less than 20%, while the liquid media had a higher percentage of ca. 50% after four successive transfers. This difference implies that a significant amount of fixed carbon by AOM is used to cross-feed and support the growth of cocultured heterotrophs [105] more on floating filters than liquid media (Supplementary Table S13). To identify the key ASV differentiating the nitrifier communities between floating filters and liquid cultures, an indicator species analysis was performed at the ASV level. Among the AOA ASVs, only ASV_4 was significantly more abundant on floating filters (IndVal: 0.6, p < 0.05) (Table 1). This ASV is closely related to members of the clade “Ca. Nitrosocosmicus” of the Nitrososphaerales lineage, with >99.9% 16S rRNA sequence similarity (Fig. 6A). In contrast, ASVs (ASV_43 and ASV_128) within the Nitrosopumilales clades did not appear to thrive on floating filters (Fig. 6B and C), differing from their growth in liquid media (Fig. 6D and E). This is consistent with our previous observation of the weak growth of “N. chungbukensis” MY2 cells on floating filters (refer to Fig. 1C and Fig. S4). The ASV_11 shows 99.19% 16S rRNA sequence similarity with Nitrospira moscoviensis BL23, indicating it might be the primary nitrite-oxidizing bacteria in floating filters and liquid cultures. The taxonomy of all nitrifiers’ ASVs is detailed in Supplementary Table S14.
A phylogenetic tree of AOA ASVs based on 16S rRNA gene sequences. Representative 16S rRNA sequences of AOA were selected from the NCBI databases. A maximum likelihood tree (A) was inferred with IQ-TREE (IQ-TREE options: -B 1000 -m MFP) using aligned sequences. Bootstrap values ≥70% based on 1000 replications are indicated. The scale bar represents a 0.1 change per nucleotide position. ASVs obtained in this work are indicated. Charts (B, C) and (D, E) represent the composition of nitrifiers of two biological replicates of floating filters and liquid cultures, respectively. The ASV_11 is affiliated with the genus Nitrospira. Details of all nitrifiers’ ASVs taxonomy are provided in Supplementary Table S14.
Indicator species analysis showing AOA ASVs on floating filters and liquid media.
| IndVal | |||||
|---|---|---|---|---|---|
| ASV_ID | Genus | Group | Floating filter culture | Liquid culture | P |
| ASV_4 | “Ca. Nitrosocosmicus” | Floating filters | 0.75 | 0.25 | .001 |
| ASV_43 | “Ca. Nitrosarchaeum” | Liquid media | 0.35 | 0.65 | .022 |
| ASV_128 | “Ca. Nitrosotenuis” | Liquid media | 0.03 | 0.97 | .001 |
| IndVal | |||||
|---|---|---|---|---|---|
| ASV_ID | Genus | Group | Floating filter culture | Liquid culture | P |
| ASV_4 | “Ca. Nitrosocosmicus” | Floating filters | 0.75 | 0.25 | .001 |
| ASV_43 | “Ca. Nitrosarchaeum” | Liquid media | 0.35 | 0.65 | .022 |
| ASV_128 | “Ca. Nitrosotenuis” | Liquid media | 0.03 | 0.97 | .001 |
AOA ASVs with significantly different relative abundances on floating filters and liquid media (IndVal >0.6, P < .05). The phylogenetic position of the ASVs is shown in Fig. 6A.
Indicator species analysis showing AOA ASVs on floating filters and liquid media.
| IndVal | |||||
|---|---|---|---|---|---|
| ASV_ID | Genus | Group | Floating filter culture | Liquid culture | P |
| ASV_4 | “Ca. Nitrosocosmicus” | Floating filters | 0.75 | 0.25 | .001 |
| ASV_43 | “Ca. Nitrosarchaeum” | Liquid media | 0.35 | 0.65 | .022 |
| ASV_128 | “Ca. Nitrosotenuis” | Liquid media | 0.03 | 0.97 | .001 |
| IndVal | |||||
|---|---|---|---|---|---|
| ASV_ID | Genus | Group | Floating filter culture | Liquid culture | P |
| ASV_4 | “Ca. Nitrosocosmicus” | Floating filters | 0.75 | 0.25 | .001 |
| ASV_43 | “Ca. Nitrosarchaeum” | Liquid media | 0.35 | 0.65 | .022 |
| ASV_128 | “Ca. Nitrosotenuis” | Liquid media | 0.03 | 0.97 | .001 |
AOA ASVs with significantly different relative abundances on floating filters and liquid media (IndVal >0.6, P < .05). The phylogenetic position of the ASVs is shown in Fig. 6A.
Unlike most other catalase-negative AOA, members of “Ca. Nitrosocosmicus” harbor a gene encoding manganese catalase [17, 19], which may confer their resistance to H2O2-induced oxidative stress while growing on floating filters. In addition, the Kdp system in the “Ca. Nitrosocosmicus” clade [102, 106] and their resilience to high saline conditions [107] further support their potential adaptation to soil surfaces. Together, our results indicate that the “Ca. Nitrosocosmicus” clade might be adapted to air-exposed soil surfaces, and the floating filter cultivation technique can be exploited to obtain novel soil-dwelling nitrifiers.
Conclusion
In this study, we demonstrated the ability of a soil ammonia-oxidizing archaeon, N. viennensis EN76, to thrive on air-exposed solid surfaces and compared its growth with that of a soil ammonia-oxidizing bacterium, N. europaea ATCC 19718. The physiological and transcriptional responses of these microorganisms highlight that the physiology of N. viennensis EN76 grown on solid surfaces (floating filter-grown cells) is notably different from the control culture in liquid media. N. viennensis EN76 exhibited significant upregulation of genes involved in the cell wall and EPS biosynthesis, H2O2-induced oxidative stress response, and ammonia oxidation when cultivated on floating filters. These adaptations are likely crucial for its survival and functionality in air-exposed solid surfaces.
Ecologically, this study underscores the importance of surface-attached growth for soil AOA. Given that soil microorganisms predominantly exist as microcolonies or biofilms within pores of soil aggregates, the ability of AOA to thrive in such conditions is critical for their role in nitrogen cycling. Thus, the floating filter cultivation approach has proven to be a valuable tool for studying the ecophysiology of soil AOA, providing insights that are more representative of their natural habitats than traditional liquid culture methods. Furthermore, the observed distinct soil AOA communities dominated by catalase-containing AOA enriched on filters compared to liquid culture suggest that the floating filter cultivation technique can indeed potentially cultivate soil surface-adapted nitrifiers. This approach presents an exciting opportunity for further exploration of the function, activity, and diversity of AOA communities in various soil environments. Overall, this study sheds light on the adaptive mechanisms governing AOA growth on air-exposed solid surfaces and its ecological relevance, paving the way for more accurate and ecologically valid investigations of the soil nitrification process.
Author contributions
S.-K.R., C.A., and J.-H.G. designed the research. C.A. and U.-J.L. processed the soil samples. C.A., J.-H.G., and M.-Y.J. performed the experiments. C.A., J.-H.G., and S.-K.R. analyzed the data. C.A. and S.-K.R. wrote the first draft of the manuscript, and J.-H.G., S.I.A., M.-Y.J., and W.P. contributed to subsequent revisions. All authors approved the final version of the manuscript.
Conflicts of interest
The authors declare no competing interests.
Funding
This work was supported by the NRF (National Research Foundation of Korea) grant funded by the Korean government (Ministry of Science and ICT) (2021R1A2C3004015), Basic Science Research Program through NRF funded by the Ministry of Education (2020R1A6A1A06046235), and Korea Institute of Marine Science & Technology Promotion (KIMST) funded by the Ministry of Oceans and Fisheries (RS-2024-00436293) J.-H.G. was supported by the NRF grant funded by the Korean government (Ministry of Science and ICT) (RS-2023-00213601). M.-Y.J. was supported by the NRF grant funded by the Korean government (Ministry of Science and ICT) (2021R1C1C1008303 and 2022R1A4A503144711).
Data availability
The whole-transcriptome data and 16S rRNA gene amplicon sequencing data generated in this study have been deposited in the NCBI BioProject database under the accession project number (PRJNA1131856) and (PRJNA1131940), respectively.
