Ecophysiological Interaction between Nitrifying Bacteria and Heterotrophic Bacteria in Autotrophic Nitrifying Biofilms as Determined by Microautoradiography-Fluorescence In Situ Hybridization

Biofilm samples.Autotrophic nitrifying biofilms were cultured with synthetic medium in a partially submerged rotating disk reactor consisting of five polymethylmethacrylate disks. The autotrophic nitrifying biofilms were first cultured with primary settling tank effluent from the Soseigawa municipal wastewater treatment plant (Sapporo, Japan) for 2 to 3 days and then were cultured with synthetic nutrient medium that contained no organic matter. The nutrient medium was composed of NH₄Cl (3.6 mM), NaHCO₃ (17.8 mM), K₂HPO₄ (0.4 mM), MgSO₄ · 7H₂O (0.41 mM), and NaCl (1.25 mM), and the pH was 7.8 ± 0.2. Distilled water was used to dilute the medium, which contained no detectable dissolved organic carbon. The reactor volume was 1,400 cm³. The total biofilm area was 2,545 cm². The temperature was maintained at 25°C. The disk rotational speed of the 18-cm-diameter disk was fixed at 14 rpm. The hydraulic retention time of the reactor was 6 h. Biofilm samples were obtained from the rotating disk reactor after more than 3 months of operation.

DNA extraction and PCR amplification.DNA was extracted from a biofilm sample (approximately 0.2 ml) with a Fast DNA spin kit (Bio 101, Qbiogene Inc., Carlsbad, Calif.), as described in the manufacturer's instructions. 16S rRNA gene fragments from isolated total DNA of the biofilm sample were amplified with Taq DNA polymerase (TaKaRa Bio Inc., Ohtsu, Japan) by using bacterial primer sets 11f (20) and 1492r (46). The conditions used for the PCR were as follows: 5 min of initial denaturation at 94°C and 20 cycles of 30 s at 94°C, 30 s at 50°C, and 1 min at 72°C. Final extension was carried out for 4 min at 72°C. The PCR products were electrophoresed on a 1% (wt/vol) agarose gel.

Cloning and sequencing of the 16S rRNA gene and phylogenetic analysis.The purified PCR products were ligated into the pGEM-T vector cloning system (Promega, Tokyo, Japan). The ligated products were transformed into Escherichia coli JM109 competent cells (Promega). Plasmids were extracted from the cloned cells and purified with the Wizard Plus Minipreps DNA purification system (Promega). Nucleotide sequencing was performed with an automatic sequencer (Prism 310 genetic analyzer; Applied Biosystems). All sequences were checked for chimeric artifacts by the Chimera Check program from the Ribosomal Database Project (22). Partial sequences (approximately 500 bp) were compared with similar sequences of the reference organisms by a BLAST search (2). Sequences with 97% or greater similarity were grouped into operational taxonomic units (OTUs) by the Similarity Matrix program from the Ribosomal Database Project (22). Nearly complete sequencing of the 16S rRNA gene of each representative OTU was performed, and the sequences were aligned with the CLUSTAL W package (44). A phylogenetic tree was constructed by the neighbor-joining method (42). Bootstrap resampling analysis of 100 replicates was performed to estimate the confidence of tree topologies.

Incubation with radioactive substrates.The following inorganic and organic substrates, labeled with radioisotopes, were used: (i) sodium [¹⁴C]bicarbonate (specific activity, 58 mCi mmol⁻¹), as a carbon source for nitrifiers; (ii) [1-¹⁴C]acetic acid (sodium salt; specific activity, 61 mCi mmol⁻¹), as a low-molecular-weight organic substrate produced through decomposition of complex organic compounds; (iii) an l-amino acid mixture uniformly labeled with ¹⁴C (specific activity, 50 mCi mmol⁻¹), as a low-molecular-weight organic substrate produced through decomposition of proteins; and (iv) N-acetyl-d-[1-¹⁴C]glucosamine ([¹⁴C]NAG) (specific activity, 57 mCi mmol⁻¹), as a major constituent of the bacterial cell wall. Radioactive chemicals were purchased from Amersham Biosciences (Little Chalfont, United Kingdom) and ICN Biomedical Inc. (Irvine, Calif.).

Four different experiments with four different types of radioactively labeled substrate were conducted (Table 1). The homogenized biofilm samples were diluted to obtain a final concentration of 3 g of volatile suspended solids per liter with a basal medium (0.4 mM K₂HPO₄, 0.41 mM MgSO₄ · 7H₂O, 1.25 mM NaCl; pH 8.0). Although the in situ spatial organization of the microbial community was disturbed, the homogenization was essential for the quantitative analysis described below. For each experiment, 1.9-ml portions of a diluted biofilm sample were transferred to 10-ml serum bottles. The cultures were supplemented with unlabeled ammonium (1 mM) and bicarbonate (0.5 mM) and with radioactively labeled inorganic and organic substrates (final radioactivity, 10 μCi per 2-ml culture). When radioactively labeled bicarbonate was used, unlabeled bicarbonate was not added. The final concentrations of radioactively labeled acetic acid, bicarbonate, NAG, and amino acids were 82, 86, 88, and 100 μM, respectively. Unlabeled acetic acid, NAG, and amino acids were not added in any experiment. The final total volume of each culture was 2 ml. The vials were sealed with gas-tight rubber stoppers and aerobically incubated for 4 h with shaking at 100 rpm at 25°C. The incubation time, 4 h, was determined in preliminary experiments to minimize changes in the microbial community structure and cascade utilization of ¹⁴CO₂ derived from oxidation of radiolabeled organic substrates during the incubation. Biofilm samples pasteurized at 70°C for 15 min were incubated in the same way with the radioactive and nonradioactive substrates in experiments performed at the same time as experiments 1 to 4 as a control to test for possible adsorption phenomena and chemography.

Liquid scintillation counting.The uptake of radioactive substrates was confirmed by liquid scintillation counting in all experiments before analysis by MAR-FISH. The ¹⁴C content was measured directly in a culture sample (biomass plus culture medium) and washed biomass that was prepared by centrifugation (10,000 × g for 8 min) of a 1-ml culture sample and then resuspension in 1 ml of tap water. Each 0.1-ml aliquot was added to 3 ml of scintillation liquid (Ultima Gold XR; Packard BioScience Co., Meriden, Conn.). After the samples were thoroughly mixed and stored at room temperature for 3 h, the radioactivity was determined with an Aloka model LSC-1000 liquid scintillation counter used as recommended by the manufacturer. To evaluate the uptake of radioactive substrates, the percentage of radioactive substrate incorporated into the biomass was calculated.

Sample fixation and washing.After incubation with radioactive and nonradioactive substrates, the samples were fixed for 3 h at 4°C by adding 2 ml of 8% paraformaldehyde, which resulted a final concentration of 4%. Subsequently, the samples were centrifuged at 10,000 × g for 8 min and washed three times with 2 ml of phosphate-buffered saline (10 mM sodium phosphate buffer, 130 mM sodium chloride; pH 7.2) to remove excess soluble radioactive substrates. After the fixation and washing steps, the samples were spotted on a gelatin-coated cover glass (16).

Oligonucleotide probes and in situ hybridization.The 16S and 23S rRNA-targeted oligonucleotide probes used in this study are shown in Table 2. The probes were labeled with fluorescein isothiocyanate (FITC), tetramethylrhodamine 5-isothiocyanate (TRITC), or the sulfoindocyanine dye Cy5. Some probes were designed in this study by using the PROBE_DESIGN tool of the ARB software package (http://www.arb-home.de/ ) according to the current version (June 2002) of the 16S rRNA sequence database. The specificity of the probes was checked against the ARB database, and the optimal hybridization conditions were experimentally determined (Table 2). All probes designed in this study were named in accordance with the standard introduced by Alm et al. (1), while the names of previously described probes were not changed to avoid confusion. Dehydration and FISH were performed by using the procedure described by Amann (4) and Okabe et al. (34). The hybridization conditions used for FISH are shown in Table 2. Simultaneous hybridizations with the probes requiring different stringency conditions were performed by using a successive hybridization procedure; hybridization with the probe requiring higher stringency was performed first, and then hybridization with the probe requiring lower stringency was performed.

The microbial community composition in the autotrophic nitrifying biofilm was quantitatively analyzed by using group- and subclass-specific probes listed in Table 2. For determination of microbial compositions, the surface fraction of the specific-probe-hybridized cell area and the EUB338 probe-hybridized cells (total biomass) was determined after simultaneous in situ hybridization with various probe sets. The average surface fraction was determined from at least 10 representative laser-scanning microscopy (LSM) projection images of each cross section of the biofilm samples by using image analysis software provided by Zeiss (34).

Autoradiographic procedure.MAR was performed directly on the cover glasses as described by Lee et al. (21). After the FISH procedure, autoradiographic liquid film emulsion (LM-1; Amersham Biosciences) was used. The optimal exposure time was adjusted to between 1 and 3 days, depending on the percentage of radioactive substrates incorporated into the biomass.

Microscopy and enumeration by MAR-FISH.A model LSM510 confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany) equipped with an Ar ion laser (458 and 488 nm) and two He-Ne ion lasers (543 and 633 nm) was used. Formation of silver grains in the autoradiographic film was observed by using the transmission mode of the system. All images were combined and processed with the standard software package provided with the LSM510 microscope. Processed images were printed by using the Photoshop 5.0 software package (Adobe Systems Inc., Mountain View, Calif.).

A MAR-positive cell was defined as a cell that was covered with more than five silver grains. The numbers of MAR-positive cells and total probe-hybridized cells were determined by directly counting a minimum of 500 silver-grain-covered cells in randomly chosen microscopic fields.

Nucleotide sequence accession numbers.The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA gene sequences of 13 clones used for the phylogenetic tree analysis are AB117705 to AB117717 .

Phylogenetic analysis.One hundred clones were randomly selected, approximately 500 bp was analyzed, and the clones were grouped into 13 OTUs on the basis of more than 97% sequence similarity within an OTU. A nearly complete 16S rRNA gene sequence of a representative clone of each OTU was analyzed, and a phylogenetic tree was generated (Fig. 1). No chimeric sequences were observed. The similarities, the closest relatives in the database, and the detection frequencies of the representative clones are shown in Table 3. The clone distribution of the clone library was as follows: genus Nitrospira, 39 of 100 clones; genus Nitrosomonas of the β-subclass ammonia-oxidizing bacteria (AOB), 25 of 100 clones; green nonsulfur bacteria (GNSB), 14 of 100 clones; members of the γ subclass of the Proteobacteria (γ-Proteobacteria), 9 of 100 clones; Cytophaga-Flavobacterium-Bacteroides (CFB) group, 9 of 100 clones; α-Proteobacteria, 3 of 100 clones; and member of the phylum Verrucomicrobia, 1 of 100 clones. Therefore, the clones affiliated with nitrifying bacteria (AOB and nitrite-oxidizing bacteria [NOB]) accounted for 64% of the clones analyzed. The sequence of OTU 10 was affiliated with the Nitrospira sp. belonging to subdivision I of the genus Nitrospira, but the level of sequence similarity to the previously isolated organism Nitrospira moscoviensis belonging to subdivision II of the genus Nitrospira (10) was only 93%. This indicated that the Nitrospira-like bacteria detected in the biofilm are most likely members of a novel species of Nitrospira. None of the clone sequences affiliated with the α-subclass NOB (i.e., the genus Nitrobacter) were detected in the biofilm. Almost all the clones belonging to heterotrophic bacterial groups exhibited less than 96% similarity to previously identified bacteria or environmental clones that have not been isolated yet in the database (Fig. 1).

• Open in new tab
• Download powerpoint

• Open in new tab
• Download powerpoint

FIG. 1.

Phylogenetic trees for α-, β-, and γ-Proteobacteria (A) and Nitrospira, Verrucomicrobia, GNSB, and the CFB group (B), showing the positions of the clones obtained from the autotrophic nitrifying biofilm. The tree was generated by using 1,309 bp (A) and 1,086 bp (B) of the 16S rRNA genes and the neighbor-joining method. Scale bar = 5% sequence divergence. The values at the nodes are bootstrap values (100× resampling analysis). The Methanobacterium congolense sequence served as the outgroup for rooting the tree. The numbers in parentheses indicate the frequencies of appearance of identical clones in the clones analyzed.

In situ spatial organization.The in situ spatial organization of heterotrophic bacteria and nitrifying bacteria was visualized by FISH by using newly designed probes and the previously described group-specific probes (Table 2). The microorganisms represented by OTUs 5 and 6 could be detected with the ALF1b probe. Large rod-shaped bacterial cells occurring singly were also detected with the ALF1b probe. These cells were evenly distributed throughout the biofilm, and some of them were present in the nitrifier clusters (Fig. 2A). The microorganisms represented by OTUs 3 and 4 could be detected with the GAM42a probe. Small, long, rod-shaped bacterial cells were found around the nitrifier clusters throughout the biofilm with the GAM42a probe (Fig. 2B). Probes S-*-CFB-0655-a-A-18 (specific for OTU 8) and S-*-CFB-0730-a-A-18 (specific for OTU 9) were designed in this study for the CFB group because previously described probes CF319a/b and CFB560 (36) did not cover all clones affiliated with the members of the CFB group detected in this study. By using probes CF319a/b, S-*-CFB-0655-a-A-18, and S-*-CFB-0730-a-A-18, thin filamentous bacteria were detected around the nitrifier clusters mainly in the surface of the biofilm (Fig. 2C). The S-*-GNS-0667-a-A-18 probe was designed to cover both OTUs 11 and 12 belonging to subdivision I of the GNSB. The filamentous morphotype (thickness, ca. 0.5 to 2.0 μm), which occurred mainly around the nitrifier clusters, was detected with probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223 (Fig. 2D). OTU 13, which is related to Opitutus sp. in the phylum Verrucomicrobia (8, 17), was not clearly detected with probe S-*-Opitu-0774-a-A-18 due to the small cell size, low fluorescence signal intensity, and low abundance.

• Open in new tab
• Download powerpoint

FIG. 2.

Confocal laser scanning microscope images showing the in situ spatial organization of nitrifying bacteria (AOB plus NOB) and heterotrophic bacteria in the autotrophic nitrifying biofilm. (A) FISH with TRITC-labeled probes Nso190 and Ntspa1026 (specific for AOB and Nitrospira, respectively) and FITC-labeled probe ALF1b (specific for α-Proteobacteria). (B) FISH with TRITC-labeled probes BET42a and Ntspa1026 (specific for β-Proteobacteria and Nitrospira, respectively) and FITC-labeled probe GAM42a (specific for γ-Proteobacteria). (C) FISH with TRITC-labeled probes Nso190 and Ntspa1026 (specific for AOB and Nitrospira, respectively) and FITC-labeled probes CF319a/b, S-*-CFB-0665-a-A-18, and S-*-CFB-0730-a-A-18 (specific for the CFB group). (D) FISH with TRITC-labeled probes Nso190 and Ntspa1026 (specific for AOB and Nitrospira, respectively) and FITC-labeled probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223 (specific for GNSB). Bars = 10 μm. The surface of the biofilm cannot be seen in the photographs.

Community composition.The microbial community composition in the autotrophic nitrifying biofilm was analyzed by FISH with various sets of probes (Table 2 and Fig. 3). Consistent with the results of the 16S rRNA gene sequence analysis, members of the Nitrospira phylum (NOB) and AOB accounted for 28 and 22% of the organisms, respectively. On the other hand, members of the α-proteobacterial, γ-proteobacterial, GNSB, and CFB groups accounted for 23, 13, 9, and 2%, respectively; these results were slightly different from the results of the 16S rRNA gene sequence analysis. The ratio of total nitrifiers (AOB plus NOB) to all heterotrophs was approximately 1:1.

• Open in new tab
• Download powerpoint

FIG. 3.

Microbial community composition for the autotrophic nitrifying biofilm as determined quantitatively by FISH. The number of cells that hybridized with a probe in each microscopic field was expressed as a percentage of the total surface area of the bacteria that hybridized with the EUB338 probe. Unidentified refers to the surface area of cells detectable with probe EUB338 but not with any other specific probe.

Substrate uptake patterns for different phylogenetic groups.The percentages of uptake (the amount of the total radioactivity incorporated in cell mass) of [¹⁴C]bicarbonate, [¹⁴C]acetic acid, ¹⁴C-amino acids, and [¹⁴C]NAG after 4 h of incubation were 7, 13, 6, and 1%, respectively. The differences in uptake of substrates may be attributed to the different microbial population sizes and the substrate uptake rate of each bacterial group. The uptake of all of the radioactive substrates was not significant (less than 0.08%) when the samples were pasteurized as a control. Consequently, no silver-grain-coated cells were observed. These results indicated that the radioactive substrates were truly incorporated by the cells rather than adsorbed on the cell surface.

The patterns of uptake of organic compounds were very different for the different phylogenetic groups. The nitrifying bacteria (AOB plus NOB) took up only [¹⁴C]bicarbonate, underlining the autotrophic growth of these organisms (Fig. 4A). No MAR-positive nitrifiers were detected in the experiment with the radiolabeled organic substrates due to the high background of [¹²C]bicarbonate, which diluted ¹⁴CO₂ derived from oxidation of radiolabeled organic substrates during 4 h of incubation. No uptake of [¹⁴C]bicarbonate by heterotrophic bacterial groups was observed after 4 h of incubation (Table 4). α-Proteobacteria that hybridized with probe ALF1b and γ-Proteobacteria that hybridized with probe GAM42a accounted for 35 and 42% of the total [¹⁴C]acetic acid-utilizing bacteria, respectively (Fig. 5A). ¹⁴C-amino acids were utilized by all heterotrophic bacterial groups (Fig. 4C and Table 4). The relative contributions to ¹⁴C-amino acid utilization reflected the community composition of heterotrophic bacteria. The GNSB that hybridized with probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223 took up ¹⁴C-amino acids and [¹⁴C]NAG (Fig. 4C and D) and accounted for 9% of the ¹⁴C-amino acid-utilizing bacterial population and 27% of the [¹⁴C]NAG-utilizing bacterial population (Fig. 5A). Interestingly, the members of the CFB group that hybridized with the CF319a/b, S-*-CFB-0665-a-A-18, and S-*-CFB-0730-a-A-18 probes were a numerically dominant group of [¹⁴C]NAG-utilizing bacteria (accounting for 64% of the bacteria), even though their abundance in the biofilm was only 2% (Fig. 5A). More than 75% of the CFB group bacteria that hybridized with the specific probes took up ¹⁴C-amino acids and [¹⁴C]NAG, demonstrating the high substrate specificity of the CFB organisms for amino acids and NAG (Fig. 5B). The unidentified group utilized all organic substrates used in this study (Fig. 5A). Although its abundance in the biofilm was relatively low (ca. 3% [Fig. 3]), the unidentified group could be an ecophysiologically important bacterial group for carbon metabolism in the biofilm. However, the phylogenetic identity of the unidentified group was not clear in this study.

• Open in new tab
• Download powerpoint

FIG. 4.

Combined MAR-FISH images of homogenized autotrophic nitrifying biofilm samples. (A) Uptake of [¹⁴C]bicarbonate by nitrifying bacteria when NH₄⁺ was used as the sole electron donor. In situ hybridization was performed with FITC-labeled probe Ntspa1026 (green), which was used simultaneously with TRITC-labeled probe Nso190 (red). (B) Uptake of [¹⁴C]acetic acid by members of the γ-Proteobacteria group. In situ hybridization was performed with FITC-labeled probe EUB338 (green) and TRITC-labeled probe GAM42a (red). (C) Uptake of ¹⁴C-amino acids by the GNSB. In situ hybridization was performed with FITC-labeled probe EUB338 (green) and TRITC-labeled probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223 (red). (D) Uptake of [¹⁴C]NAG by the GNSB. In situ hybridization was performed with FITC-labeled probe EUB338 (green) and TRITC-labeled probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223 (red). Yellow signals resulted from binding of both FITC- and TRITC-labeled probes to one cell. Bars = 10 μm.

• Open in new tab
• Download powerpoint

FIG. 5.

Patterns of substrate uptake by the major phylogenetic bacterial groups identified by using group-specific probes. (A) Fractions of group-specific-probe-hybridized cells that simultaneously took up radioactive substrates (MAR-positive cells), expressed as percentages of the total bacteria that were detected with probe EUB338 and simultaneously took up radioactive substrates. (B) Fractions of probe-hybridized and MAR-positive cells, expressed as percentages of each group-specific-probe-hybridized cell count. AOB, AOB that hybridized with probe Nso190; NOB, NOB that hybridized with probe Ntspa1026; α-Proteobacteria hybridized with probe ALF1b; γ-Proteobacteria hybridized with probe GAM42a; GNSB, GNSB that hybridized with probes S-*-GNS-0667-a-A-18, GNSB-941, and CFX1223; CFB, members of the CFB group that hybridized with probes CF319a/b, S-*-CFB-0665-a-A-18, and S-*-CFB-0730-a-A-18. The error bars indicate the standard errors (n = 10).

Substrate uptake patterns of heterotrophs.The patterns of uptake of low-molecular-weight dissolved organic matter differed greatly for the different phylogenetic groups. Despite of the low abundance (ca. 2%) in the biofilm community, the members of the CFB cluster accounted for the largest fraction (ca. 64%) of the bacterial community consuming NAG. Although the α- and γ-Proteobacteria were numerically dominant bacterial groups, they did not utilize NAG at all.

These results are consistent with previous studies that demonstrated that some members of the CFB group degrade various refractory biomacromolecules, such as cellulose, chitin, DNA, lipids, and proteins, which should have been abundant in the biofilm, in which dead microorganisms were trapped (36). These organisms are also known for their ability to produce exopolysaccharide slime, suggesting that they primarily form biofilms. It is, however, likely that the abundance of members of the CFB group was potentially underestimated due to a lack of CFB-specific probes that fully cover the whole CFB division and the low fluorescence intensity (i.e., low 16S rRNA content) in the biofilm.

Despite the low abundance (ca. 9%), the members of the GNSB cluster also played an important role in degradation of [¹⁴C]NAG, accounting for 27% of the [¹⁴C]NAG-consuming bacterial population (Fig. 5A). Based on 16S rRNA gene analysis, all our clone sequences were affiliated with GNSB subdivision I. Subdivision I contains the most diverse environmental clones among the four subdivisions; these clones were derived from various oxic and anoxic environments (15). It could be speculated that the GNSB could utilize a yeast extract-like substrate that includes amino acids released by the autolysis of heterotrophic bacteria in the biofilm, as suggested by Sekiguchi et al. (43).

The α- and γ-proteobacterial populations were numerically dominant heterotrophic populations in the biofilm. High levels of α-Proteobacteria in an autotrophic nitrifying biofilm and in oligotrophic river biofilms have been reported previously (25, 31). It is probable that α- and γ-Proteobacteria primarily utilize low-molecular-weight fatty acids produced from degradation of SMP produced by nitrifiers. Some α-Proteobacteria strains also could take up NAG (9, 40). The most frequently detected clone sequences were affiliated with the Hyphomonas cluster of α-Proteobacteria (98% similarity) and the Xanthomonas cluster of γ-Proteobacteria (91 to 95% similarity) (Fig. 1A). Hyphomonas spp. catabolize proteins, peptides, and/or amino acids for energy and growth and form biofilms by production of extracellular polymeric substances (45). Xanthomonas spp. are able to utilize l-arabinose that is a constituent of the bacterial cell wall under aerobic and microaerophilic conditions (7).

Carbon metabolism in the autotrophic nitrifying biofilm.It is thought that the primary degraders directly colonize or occur around the nitrifier clusters that liberate (excrete) SMP. They then recruit other species to form metabolically structured and functionally integrated biofilm communities that ensure maximum utilization of metabolites (i.e., SMP) of nitrifiers. The in situ spatial organization of the nitrifier clusters and heterotrophs in the biofilm was visualized by FISH (Fig. 2). In this study, GNSB were detected mainly around nitrifier clusters in a biofilm for the first time (6, 19). We also found that there was a tendency for members of the CFB group, GNSB, and γ-Proteobacteria to be more closely associated with the nitrifier clusters. However, we could not determine whether there was a clear relationship between the spatial organization of different groups of heterotrophs and their physiological roles in carbon metabolism in the biofilm.

The ratio of nitrifiers (AOB plus NOB) to heterotrophs was approximately 1:1 in the biofilm. According to the kinetic parameters for SMP production by nitrifiers (41), we estimated that only about 1.5 mg of chemical oxygen demand per liter of SMP could be produced from 3.6 mM NH₄⁺ in the feed. Therefore, the carbon metabolism in the biofilm must be efficient to support heterotrophic bacteria (ca. 10⁸ to 10⁹ cells cm⁻³) with this small amount of SMP. SMP have proved to be major components of organic matter in effluents from biological wastewater treatment processes (27). Our knowledge regarding SMP of nitrifiers is, however, far from complete, and much work is still required to fully understand their contribution to the growth of heterotrophic bacteria.

In conclusion, a full-cycle 16S rRNA approach followed by MAR-FISH revealed that α- and γ-Proteobacteria mainly utilized low-molecular-weight organic matter, like [¹⁴C]acetic acid and ¹⁴C-amino acids, whereas the GNSB and CFB group were specialized and decomposed [¹⁴C]NAG despite low abundance in the autotrophic nitrifying biofilm. The phylogenetically diverse and (to some extent) metabolically redundant heterotrophic bacterial community could have ensured effective utilization of SMP produced by nitrifiers, which may have created a functionally structured stable biofilm ecosystem. To determine how the radioactivity of [¹⁴C]bicarbonate incorporated by nitrifiers is transferred to major phylogenetic groups of heterotrophs, the fate of the radioisotopically labeled atom of [¹⁴C]bicarbonate should be directly traced by MAR-FISH in the future.