A moderately thermophilic ammonia-oxidizing crenarchaeote from a hot spring

Initially, the bacterial diversity of the enrichment was monitored by PCR amplification of 16S rRNA gene sequences by using the general bacterial primers 616V and 630R. A total of 42 clones were screened by restriction fragment length polymorphism (RFLP) and assigned to five different patterns; 18 clones representing all RFLP types were sequenced and phylogenetically analyzed (SI Fig. 5). All clones formed a monophyletic group within the Betaproteobacteria, but no 16S rRNA gene sequences related to known AOB or nitrite oxidizers were found. Furthermore, no amplicon was obtained with a 16S rRNA gene-based PCR assay for betaproteobacterial AOB by using primers βamoF and Nso1225R. A combination of these primers with primers targeting the domain Bacteria (616V or 630R) also did not yield any PCR product. The enrichment was also screened for amoA genes of beta- and gammaproteobacterial AOB, but although the applied assays together target all known AOB, no PCR product was observed. Identical results were observed with the addition of PCR-enhancing substances, suggesting that the negative results were not caused by PCR inhibition. This finding was further confirmed by the addition of suitable control DNA to the enrichment biomass before PCR, which always led to successful amplification (data not shown). Consistent with the PCR results, no FISH signals were detectable after hybridization of the formaldehyde-fixed biomass with a suite of AOB- and nitrite oxidizer-specific probes (SI Table 1).

Archaeal 16S rRNA genes were PCR-amplified by using two different sets of primers, cloned, RFLP-screened, and sequenced. In total, 46 clones representing all RFLP types were partly or fully sequenced. All 46 sequences were related at >99.1% similarity to each other and assigned to a monophyletic cluster within Crenarchaeota group I.1b (soil group; Fig. 1) that also included partial 16S rRNA gene sequences recently retrieved from a subsurface radioactive thermal spring (20) and a Wisconsin soil (28). In addition, crenarchaeotal amo genes were PCR-amplified from the enrichment, and 19 and 7 clones were sequenced for amoA and amoB, respectively. All cloned amoA and amoB sequences showed similarities of >99.2% and >99.3% to each other, respectively. No amplification product for amoC could be obtained. Consistent with the 16S rRNA tree topology, the AmoA and AmoB sequences retrieved from the enrichment formed a monophyletic branch within the crenarchaeotal I.1b (soil) cluster (Fig. 2A and SI Fig. 6).

To directly link the retrieved crenarchaeotal 16S rRNA gene sequences with single microbial cells in the enrichment, catalyzed reporter deposition (CARD)–FISH with the newly designed clone-specific probe RHGA702 was applied. This probe hybridized exclusively to small cocci of 0.9 ± 0.3 μm (n = 22) in diameter, occurring in irregular-shaped microcolonies of 10–50 μm in diameter (Fig. 3). Not all cells of such microcolonies showed a signal after hybridization. Identical results were observed in parallel control experiments with the crenarchaeotal probe Cren512 and the general archaeal probe Arch915. The relative abundance of the crenarchaeotes in the enrichment was determined by an indirect CARD-FISH assay to be 50.4% ± 12.5 (SD) (for details, see SI Results).

Transcription of crenarchaeotal amoA was demonstrated with biomass from the enrichment (grown in batch culture with 2.0 mM ammonium in the medium) via RT-PCR by using primers specifically targeting the obtained amoA sequences (amo16F/amo586R) (Fig. 2B). Cloning and sequencing of the RT-PCR product resulted in amoA sequences that were identical to amoA sequences obtained by PCR amplification from the enrichment.

Ammonia-oxidizing activity of the detected crenarchaeote was demonstrated by CARD-FISH/microautoradiography (MAR). The enriched biomass was preincubated at 46°C for 3 h with different ammonium concentrations without addition of labeled bicarbonate to allow adaptation of the cells to the incubation conditions. After this step, the ammonium concentration was determined, radioactive bicarbonate was added as sole external carbon source, and the biomass was incubated at 46°C for 12 h. For all ammonium concentrations measured (0.14 mM, 0.79 mM, and 3.08 mM, after preincubation) MAR signals coincided with the AOA microcolonies, but bicarbonate fixation was much lower in the presence of 3 mM ammonium than at lower ammonium concentrations (Fig. 3). For all experiments, chemical analyses showed that 12–22% of the ammonium was removed during the 12-h incubation step, and no cells other than AOA identified by the specific FISH probe showed a positive MAR signal (SI Fig. 7). It should, however, be noted that because the maximum exposure time examined was 15 days, unknown (bacterial) ammonia oxidizers with a very weak activity might have been overlooked. Neither CARD-FISH nor MAR signals were detected for AOA microcolonies in control experiments without added ammonium or with UV-sterilized biomass (data not shown). The absence of CARD-FISH signals without ammonium in the medium suggests that the AOA either reduce their cellular ribosome content in response to starvation and/or that, under these conditions, changes in the cell envelope composition occurred, which negatively affected probe permeability. Addition of 100 μM of the inhibitor allyl-thiourea (AlTU) in the presence of 0.76 mM ammonium resulted in strongly decreased MAR signals of the AOA microcolonies (Fig. 3C).

The concentration of ammonium was measured by fluorescence detection according to Corbin (55) after precolumn-derivatization with OPA reagent and HPLC separation. Nitrite and nitrate concentrations were determined quantitatively by UV detection after ion-pair chromatography on a Hypersil ODS C18 column (56). The oxygen concentration in enrichment flasks was determined with an electrode (OXI 96; Nova Analytics).

PCR amplifications of archaeal and bacterial 16S rRNA and amo genes were performed directly from harvested cells (diluted in 0.9% NaCl) from an ammonia-oxidizing enrichment from the Garga hot spring in the Buryat Republic, Russia (27), by using primers and annealing temperatures as listed in SI Table 2. At the time of analyses, this enrichment had been maintained for 6 years but had not been screened for ammonia oxidizers with nucleic acid-based methods. All PCRs started with an initial heating step at 95°C for 5–7 min to induce cell lysis. PCR-enhancing substances (betaine, bovine serum albumin, dimethyl sulfoxide, or formamide) were added according to the literature (57, 58). Cloning and sequencing were done as described in ref. 59. The diversity of sequences within the clone libraries was screened with RFLP, by using the enzyme MspI (Fermentas Life Sciences Inc.). From each resulting RFLP pattern, at least one representative clone was sequenced. Sequencing was performed by using the BigDye Terminator Cycle Sequencing Kit v3.1 and the ABI 3130xl Sequencer (Applied Biosystems) following the manufacturer's protocols. Obtained sequences were phylogenetically analyzed by using the software programs ARB (60) and PHYLIP (61) with comprehensive databases that contained 16S rRNA and all available archaeal amo sequences, respectively.

Cells were harvested from the enrichment by centrifugation and were stored in RNAlater (Ambion) at 4°C until further processing. Total RNA was isolated by using TRIzol (Invitrogen) according to the manufacturer's instructions with the following modifications: RNA was precipitated overnight at −20°C in the presence of 5 μg of glycogen (Sigma). After incubation with DNase (Sigma), reverse transcription was carried out by using the Revert Aid First Strand cDNA Kit (Fermentas Inc.) according to the manufacturer's instructions with the clone-specific primer amo586R (5′-AGC AAT GGG AAC TGA CAG-3′), which had been designed by using ARB. cDNA was amplified by using clone-specific primers amo16F (5′-ACG CAC AAC GCA CTA CTT-3′) and amo586R (length of amplificate: 570 bp) with thermal cycling as follows: initial denaturation at 94°C for 4 min followed by 40 cycles of denaturation at 94°C for 40 s, annealing at 60°C for 30 s, and elongation at 72°C for 40 s, followed by a final elongation step at 72°C for 10 min. Specificity of amplification was confirmed by cloning and sequencing of the amplification product. Absence of contaminating DNA in the RNA extract was demonstrated by performing the aforementioned PCR without the initial reverse transcription step.

To visualize AOB and nitrite oxidizers in the enrichment, FISH was performed as described by Daims et al. (62) by using probes listed in SI Table 1. No unspecific labeling of cells was observed in any sample by using the nonsense probe. After FISH, cells were stained at 4°C with the DNA-binding dye at 1 μg ml⁻¹. Microscopic observation and documentation were accomplished by using a LSM 510 scanning confocal microscope (Zeiss) and the included software.

To visualize AOA in situ, a specific 16S rRNA-targeting FISH-probe was designed by using ARB. Probe RHGA702 (SI Table 1) targeted all crenarchaeotal 16S rRNA sequences obtained from the enrichment but has at least one central mismatch to all other sequences in our ARB database including all published sequences from Crenarchaeota. For more details regarding this probe, please refer to probeBase (63) at http://www.microbial-ecology.net/probebase. HPLC-purified and lyophilized HRP-conjugated oligonucleotides were obtained from Thermo Electron. For CARD-FISH analyses, ethanol-fixed biomass from the enrichment was immobilized on slides and dehydrated by an increasing ethanol series (50%, 80%, and 96% for 3 min each). Afterward, the slide was dipped into 0.2% agarose (in distilled water) and left to dry at 30°C. Then, permeabilization of cell walls was performed by using 10-min incubation with proteinase K (15 μg ml⁻¹; Sigma). This and all other incubation and washing steps were done at room temperature if not stated otherwise. After permeabilization, the slides were washed for 1 min in distilled water. Subsequently, slides were incubated in 0.01 M hydrogen chloride for 20 min to destroy remaining proteinase K. To inactivate endogenous peroxidases, cells were incubated in a solution of 0.15% hydrogen peroxide in methanol for 30 min. Finally, the slides were washed two times in distilled water for 1 min. The biomass was covered with 20 μl of hybridization buffer (64) at a probe concentration of 0.16 ng μl⁻¹ (65). Hybridization was performed at 46°C for 2.5 h by using the probes Arch915, Cren512, and RHGA702 (10% formamide) in separate experiments (SI Table 1). By using a “nonsense” probe (SI Table 1), no unspecific labeling of cells could be observed. After hybridization, slides were washed in prewarmed washing buffer (64) at 48°C for 15 min. Before tyramide signal amplification, the slides were dipped into distilled water (4°C) and washed for 15 min in 1× PBS. Excess liquid was removed by tapping the slides, and the slides were incubated at 46°C for 45 min with substrate mix by using 1/1,000 parts fluorescein-labeled tyramide (65). Subsequently, the slides were washed in 1× PBS for 10 min and distilled water for 1 min, respectively. Finally, the slides were DAPI-stained and evaluated microscopically as described above. The diameter of the AOA was determined quantitatively by using the digital image analysis software daime (66).

For bicarbonate uptake experiments, cells were concentrated from the enrichment via centrifugation (10 min at 13,000 × g), washed in growth medium (27) without ammonium, centrifuged again and resuspended in a small volume of this medium. Resuspended cells were kept at 4°C for no longer than 7 h before the incubation experiments. Resuspended biomass was exposed in parallel experiments to different ammonium concentrations by using a stock solution of 10 mM NH₄Cl in a total volume of 3 ml of growth medium. Two biological replicates were performed for each of these experiments. Furthermore, control experiments were performed without addition of ammonium as well as with UV-sterilized biomass to check for physiological activity without added substrate and for chemography, respectively. For all experiments, the biomass was incubated in vertically fixed vials under aerobic conditions for 3 h at 46°C and was shaken at 110 rpm in a water bath. After this preincubation step, which was included to allow the cells to re-adapt to elevated temperature conditions, a subsample was taken (to determine the ammonium concentration via ion exchange chromatography with chemically suppressed conductivity detection) and 10 μCi [¹⁴C]bicarbonate (Hanke Laboratory Products) was added to each vial. Subsequently, the biomass was incubated for 12 h under the conditions described above. In addition, 100 μM of the inhibitor AlTU (Fluka) was added after the preincubation to one of the vials containing 0.76 mM ammonium after preincubation. After incubation, biomass was fixed with EtOH or formaldehyde as described by Daims et al. (62). CARD-FISH and subsequent DAPI staining were performed as described above. The hybridized samples were dipped in preheated (48°C) LM-1 emulsion (Amersham), exposed for 12–15 days at 4°C in the dark and developed in Kodak D19 (40 g liter⁻¹ of distilled water) before microscopic examination. For each condition and biological replicate, three microautoradiography slides were analyzed and at least 10 AOA microcolonies were examined per slide.