Dynamics of a methanol-fed marine denitrifying biofilm: 1-Impact of environmental changes on the denitrification and the co-occurrence of Methylophaga nitratireducenticrescens and Hyphomicrobium nitrativorans

Seawater media

The artificial seawater (ASW) medium was composed of (for 1 liter solution): 27.5 g NaCl, 10.68 g MgCl₂*6H₂O, 2 g MgSO₄*7H₂O, 1 g KCl, 0.5 g CaCl₂, 456 µL of FeSO₄*7H₂O 4 g/L, five mL of KH₂PO₄51.2 g/L, five mL of Na₂HPO₄ 34 g/L. The Instant Ocean (IO) seawater medium was bought from Aquarium systems (Mentor, OH, USA) and dissolved at 30 g/L. Estimation of its composition and comparison with the ASW medium is provided in Table S1. Both media were supplemented with one mL of trace elements (FeSO₄*7H₂O 0.9 g/L, CuSO₄*5H₂O 0.03 g/L and MnSO₄*H₂O 0.2 g/L) and with NaNO₃ (Fisher Scientific Canada, Ottawa, ON, Canada) at various concentrations, depending on the experiment (Tables 1 and 2). The pH was adjusted (NaOH) at 8.0 before autoclaving. Filter-sterilized methanol (Fisher Scientific) was then added as a carbon source to support bacterial growth, also at various concentrations depending on the experiment (Tables 1 and 2). The sterile media were distributed (60 ml) in sterile serologic vials, which included twenty carriers (Bioflow 9 mm; Rauschert, Steinwiessen, Allemagne). Prior to use, these carriers were washed with HCl 10% (v/v) for 3 h, rinsed with water and autoclaved. The vials were purged of oxygen for 10 min with nitrogen gas (Praxair, Mississauga, ON, Canada) and sealed with sterile septum caps.

The reference biofilm cultures (Ref300N-23C)

The carriers (Bioflow 9 mm) containing the denitrifying biofilm (here named the Original Biofilm; OB) were taken from the denitrification reactor of the Montreal Biodome and frozen at −20 °C in seawater with 20% glycerol (Laurin et al., 2006) until use. Biomass from several carriers were thawed, scraped, weighted and dispersed in the ASW at 0.08 g (wet weight)/mL. The biomass (five mL/vial, 0.4 g of biofilm) was then distributed with a syringe and an 18_G1 $\frac{1}{2}$ needle into vials containing twenty free carriers and 60 mL ASW and supplemented with 300 mg NO₃⁻-N/L (21.4 mM NO₃⁻) and 0.15% (v/v) methanol (C/N = 1.5). The vials were incubated at 23 °C and shaken at 100 rpm (orbital shaker).

On average once a week, the twenty carriers were taken out of the vial, gently washed to remove the excess medium and the free bacteria, transferred into a new vial containing fresh anaerobic medium, and then incubated again under the same conditions (Fig. S1). Samples were taken each one or two days to measure the concentrations of NO₃⁻ and NO₂⁻ in the growth medium. The residual suspended biomass was taken for DNA extraction. Methanol and NaNO₃ were added when needed if NO₃⁻ was completely depleted during the week. During the fifth carrier-transfer cultures, multiple samples (500 µL) were collected at regular intervals for 1–3 days to determine the NO₃⁻ and NO₂⁻ concentrations. After this monitoring, total biomass in a whole vial was determined by protein quantification.

Short exposure of the reference biofilm cultures to physico-chemical changes

The carriers containing the reference biofilm cultures were used as starting material to test the impact on the denitrification performance of the biofilm by varying specific physico-chemical parameters (Table 1). Figure S2 describes in detail the protocol. To perform these assays, fifteen carriers from the reference biofilm cultures were distributed into vials in the prescribed conditions described in Table 1. For the assays that were performed with different pHs and temperatures and different NaCl concentrations, the vials were incubated for two days for the biofilm to adapt to the new culture conditions, and then the carriers were transferred into their respective fresh medium (Fig. S2). NO₃⁻ and NO₂⁻ concentrations in the vials were then monitored by collecting 500 µL medium samples at regular intervals for 1 to 3 days. Biomass was taken at the end of this monitoring to extract DNA and for protein quantification. All assays were performed in triplicates, except those with NO₂⁻.

Cultivation of the original biofilm to different culture conditions

The OB was taken and processed as described in The reference biofilm cultures subsection. The formulation of the ASW medium was adjusted with the prescribed NaCl, NO₃⁻, NO₂⁻ and methanol concentrations, and pHs, and the vials were incubated at prescribed temperatures (Table 2; Fig. S3). Cultivation of the OB under oxic conditions were performed as for the reference biofilm cultures but in 250-mL Erlenmeyer flasks with agitation (Table 2). Culturing the OB in the IO medium was performed as for the reference biofilm cultures (Table 2). All biofilm cultures were shaken at 100 rpm (orbital shaker), or 200 rpm for the oxic cultures. Five carrier-transfer cultures were performed, after which the NO₃⁻, NO₂⁻ and protein concentrations were determined as described in The reference biofilm cultures subsection.

Determination of the denitrification rates

Measurements of NO₃⁻ and NO₂⁻ concentrations in the biofilm cultures were performed in all assays using ion chromatography as described by Mauffrey et al. (2017). The total biomass in a vial was measured by collecting the suspended biomass and the biomass attached to the carriers (Fig. S1 to Fig. S3). The amount of protein in the biomass was determined by the Quick Start™ Bradford Protein Assay (BioRad, Mississauga, ON, Canada). The denitrification rates were calculated by the linear portion of the NO₃⁻ plus NO₂⁻ concentrations (NO_x) over time for each replicate (mM h⁻¹). The specific denitrification rates were reported as the denitrification rates divided by the quantity of biomass (mg protein) in a vial (mM h⁻¹ mg-protein⁻¹). Ammonium concentration was measured using the colorimetric method described by Mulvaney (1996).

DNA-based analyses

The DNA was extracted from the suspended biomass and the biofilm that was scraped from the carriers as previously described (Geoffroy et al., 2018). The PCR amplifications of the V3 region of the 16S ribosomal RNA (rRNA) genes for denaturating gradient gel electrophoresis (DGGE) experiments were performed as described (Lafortune et al., 2009) with the 341f and 534r primers (Table S2). Total DNA extracted from the planktonic pure cultures of strains JAM1 and NL23 was used to perform PCR amplifications of the same region of the 16S rRNA genes. The resulting strain-derived amplicons were co-migrated with the biofilm cultures-derived amplicons on DGGE gels to identify the DNA fragments associated to strain JAM1 and strain NL23 in biofilm samples.

Quantitative PCR assays (SYBR green) to measure the concentrations of M. nitratireducenticrescens strains JAM1/GP59 (narG1), M. nitratireducenticrescens strain GP59 (nirK), M. nitratireducenticrescens strain JAM1 (tagH) and H. nitrativorans strain NL23 (napA) (Table S2) in the biofilm cultures were performed as previously described (Geoffroy et al., 2018).

Statistics

Statistical significance in the denitrification rates between different types of cultures was performed with one-way ANOVA with Tukey’s Multiple Comparison Test.

Establishment of the reference biofilm cultures (Ref300N-23C)

The original biofilm (OB) of the Biodome denitrification reactor was used to inoculate vials containing new carriers to allow the development of a fresh denitrifying biofilm on the carrier surface. The vials were incubated under batch-mode anoxic conditions in a homemade ASW medium (2.75% NaCl) with 300 mg NO₃⁻-N/L (21.4 mM NO₃⁻) and 0.15% (v/v) methanol (C/N = 1.5) at 23 °C. The choice of ASW medium was based on our ability to change the composition of the medium, as opposed to the commercial seawater medium (Instant Ocean) used by the Biodome. Each week for five weeks, only the carriers were transferred in fresh medium (Fig. S1). The biofilm cultures cultivated under these conditions are referred as the reference biofilm cultures and named from here as the Ref300N-23C biofilm cultures (for 300 mg NO₃⁻-N/L, 23 °C). As its name implies, the conditions of these cultures were used as a reference to measure changes in the denitrifying activities resulting from changes in physico-chemical parameters in the culture medium.

The NO₃⁻ and NO₂⁻ concentrations were analyzed sporadically for the first four carrier-transfer cultures, and showed increasing rates of denitrifying activities with subsequent transfers. During the fifth carrier-transfer cultures, the concentrations of NO₃⁻ and NO₂⁻ were analyzed more thoroughly. NO₃⁻ was consumed in 4 to 6 h, with accumulation of NO₂⁻ that peaked at 10 mM after 4 h. NO₂⁻ was completely consumed after 12 h (Fig. 1A). The denitrification rates were calculated at 1.69 (±0.14) mM-NO_x h⁻¹. Taking into account the biomass, the specific denitrification rates corresponded to 0.0935 (±0.0036) mM-NO_x h⁻¹ mg-protein⁻¹. The production of ammonia was negative in these conditions, which rules out dissimilatory NO₃⁻ reduction to ammonium. Further carrier transfers did not improve the specific denitrification rates.

The evolution of the bacterial community of the OB under these conditions was assessed by PCR-DGGE profiles for each carrier-transfer cultures (Fig. 1B). These profiles were similar after the third carrier-transfer cultures, suggesting stabilization of the bacterial community in these cultures. The most substantial changes in these profiles compared to the OB profile is the intensity of the DGGE band corresponding to M. nitratireducenticrescens strain JAM1 that was much stronger in all carrier-transfer cultures, whereas the DGGE band corresponding to H. nitrativorans strain NL23 was no longer visible after the second carrier-transfer cultures (Fig. 1B). qPCR analysis confirmed these results (Fig. 1C). Strain JAM1 increased in concentration in the Ref300N-23C biofilm cultures from 5.6 × 10³ narG1 copies/ng DNA in the OB to 2.0 × 10⁵ narG1 copies/ng DNA in the fifth carrier-transfer cultures. The concentrations of strain NL23 however dropped by three orders of magnitude from 8.7 × 10⁴ in the OB to 3.2 ×10¹ napA copies/ng DNA in the fifth carrier-transfer cultures.

These results showed important changes in the bacterial community of OB occurred when cultivated in the ASW medium under the batch-operating mode, with the important decrease in concentration of the main denitrifying bacterium in the biofilm, H. nitrativorans strain NL23, without the loss of denitrifying activities. We discovered later that a new M. nitratireducenticrescens strain, strain GP59 with full denitrification capacity, was enriched in these cultures (Geoffroy et al., 2018). qPCR assays targeting narG1 cannot discriminate strain JAM1 from strain GP59, as this gene is identical in nucleic sequence in both strains (Geoffroy et al., 2018). Therefore, the concentrations of strain JAM1 in the biomass derived from narG1-targeted qPCR assays and illustrated in Fig. 1C reflects in fact the concentrations of both strains.

Impact of a short exposure to environmental changes on the Ref300N-23C biofilm cultures

The colonized carriers from the Ref300N-23C biofilm cultures were exposed for a short period (few hours to few days) to a range of methanol, NO₃⁻, NO₂⁻ and NaCl concentrations, and to different pHs and temperatures (Table 1, Fig. S2, Table S3). These conditions were chosen to assess the capacity range of the denitrifying activities of the biofilm (methanol and NO₃⁻ concentrations), but also its resilience to adverse conditions that could occur during the operation of a bioprocess (e.g., abrupt change of pH, temperature or salt concentration).

Increasing the concentrations of NO₃⁻ and methanol (fixed C/N at 1.5) showed increases in the denitrification rates compared to the original conditions of the Ref300N-23C biofilm cultures (300 mg-NO₃⁻-N/L with 0.15% methanol), with the highest increases observed at 600 mg-NO₃⁻-N/L with 0.3% methanol (58% increase), and at 900 mg-NO₃⁻-N/L with 0.45% methanol (56% increase) (Fig. 2). Inhibition of the denitrifying activities occurred at 3000 mg-NO₃⁻-N/L. This inhibition could have been caused by the methanol concentration (1.5%) in the medium. Variations in methanol concentrations (0% to 0.5% with fixed concentration of NO₃⁻ at 300 mg-NO₃⁻-N/L) showed no improvement of the denitrification activities compared to the original conditions (Fig. 2). As methanol is the only source of carbon in the medium, its absence (0%) resulted in very weak denitrifying activities as expected. Variations in NO₃⁻ concentrations (fixed concentration of methanol at 0.15%) showed the highest denitrification rates (74% increase) at 900 mg-NO₃⁻-N/L (Fig. 2). Denitrifying activities were observed this time at 3000 mg-NO₃⁻-N/L. However, NO₃⁻ consumption stopped after 5 days and 50% consumption. Addition of methanol (0.15%) allowed activities to resume for one day with a further 20% NO₃⁻ consumption.

Exposure of the Ref300N-23C biofilm cultures to pH 4 and 6 showed 2.7 and 2.6-fold increases of the denitrification rates, respectively, compared to the original conditions (pH 8) (Fig. 2). However, the pH at the end of the assays (10 to 12 h) increased at around 8 in the medium. At pH 10, the denitrifying activities were completely inhibited. For the temperature assays, the highest denitrification rates were recorded at 30 and 36 °C (Fig. 2), with 53% and 62% increases, respectively, compared to the original conditions (23 °C). At 5 °C, the denitrifying activities were still occurring but at 4-time lower rates than at 23 °C. At 15 °C, the denitrification rates were not significantly different from the original conditions. Exposure of the Ref300N-23C biofilm cultures for 3 days in ASW with NaCl concentrations ranging from 0% to 8% showed little effect on the denitrification rates (Fig. 2). Finally, presence of NO₂⁻ has a strong effect on the denitrification rates with 11-fold and 6.6-fold decreases with 400 mg-NO₂⁻-N/L and 200 mg-NO₂⁻-N/L, respectively in the medium, compared to the original conditions (Fig. 2). Mixture of 200 mg-NO₂⁻-N/L and 200 mg-NO₃⁻-N/L in the medium had a less pronounce effect with a 3-fold decrease of the denitrification rates. Although these last three assays with NO₂⁻ were performed with only one replicate, it revealed the tendency of NO₂⁻ to affect negatively the denitrifying capacity of the biofilm.

No significant differences in the quantity of biomass (protein amount per vial) was observed at the end of all these assays between vials. In all the tested conditions, the bacterial diversity profiles (PCR-DGGE profiles) showed no changes (Fig. 1B, lane Tr5B). The levels of M. nitratireducenticrescens strains JAM1/GP59 was around 10⁵ narG1 copies/ng DNA and of H. nitrativorans strain NL23 between 10¹ and 10⁻¹ napA copies/ng DNA.

Cultivation of the original biofilm under optimal conditions

As described above, the Ref300N-23C biofilm cultures exposed for a short period to 900 mg-NO₃⁻-N/L (0.45% methanol; C/N 1.5) or to 300 mg-NO₃⁻-N/L (0.15% methanol; C/N 1.5) at 30 and 36 °C showed between 53% to 74% increases in the denitrification rates compared to the original conditions (Fig. 2). Based on these results, we cultivated the OB under these optimal conditions in the attempt to derive a biofilm with higher denitrifying capacity. As before, the dispersed biomass of the OB was used as inoculums to colonize new carriers. Five carrier-transfer cultures were carried out in ASW medium supplemented with 900 mg-NO₃⁻-N/L and 0.45% methanol (C/N = 1.5), and incubated at either 23 °C or 30 °C (here named 900N-23C and 900N-30C, respectively), or at 300 mg-NO₃⁻-N/L and 0.15% methanol (C/N = 1.5) at 30 °C (here named 300N-30C). In parallel, the Ref300N-23C biofilm cultures were derived by the same protocol (Table 2, Fig. S3). The specific denitrification rates of these biofilm cultures were 20% to 85% higher than those found in the Ref300N-23C biofilm cultures (Table 3). Assuming methanol was not a limiting factor in these cultures (not all methanol was consumed after NO₃⁻ reduction), two-way ANOVA showed that temperature was the main factor affecting the specific denitrification rates (p < 0.001), not the concentrations of NO₃⁻ (p > 0.05).

The bacterial diversity profiles (derived by PCR-DGGE) of these four biofilm cultures were all similar (same profile illustrated in Fig. 1B, lane Tr5B). qPCR analyses showed the same trend in the dynamics of the populations of M. nitratireducenticrescens strains JAM1/GP59 and H. nitrativorans strain NL23. The concentrations of strains JAM1/GP59 increased by two orders of magnitude (1.9 to 2.3 × 10⁵ narG1 copies/ng DNA), whereas the concentrations of strain NL23 decreased by 2–3 orders of magnitude (6.1 to 18 × 10¹ napA copies/ng DNA) in the four biofilm cultures compared to the OB (Fig. 3A). After the isolation of strain GP59 and the sequencing of its genome, qPCR primers specific to strain JAM1 and to strain GP59 were developed (Geoffroy et al., 2018). The concentrations of these two strains were then measured in the OB and in the four biofilm cultures. Between 1.6 and 2.9 ×10⁵ nirK copies/ng DNA of strain GP59 were measured in these biofilm cultures, which is three orders of magnitude higher than what was measured in the OB (1.6 ×10² nirK copies/ng DNA). The concentrations of strain JAM1 in the four biofilm cultures ranged from 2.4 to 8.5  × 10² tagH copies/ng DNA, which is 3 to 10 times lower than its concentration in the OB (2.5 ×10³ tagH copies/ng DNA) (Fig. 3A).

Cultivation of the original biofilm with different NaCl concentrations

In the first set of assays, we showed that the Ref300N-23C biofilm cultures could sustain high variations of NaCl concentrations for a short period without affecting the denitrification capacity. Based on these results, we tested the capacity of the OB to acclimate to a range of NaCl concentrations. The dispersed biomass of the OB was cultivated during five carrier-transfers under anoxic conditions in modified ASW medium with NaCl concentrations ranging from 0% to 8% (Table 2, Fig. S3). The ASW used for the Ref300N-23C biofilm cultures contains 2.75% NaCl.

In biofilm cultures cultivated with low NaCl concentrations (0, 0.5 and 1.0%), the amounts of biomass developed were 1.3 to 2 times lower than in the Ref300N-23C biofilm cultures (Table 3). This lower development did not impair the specific denitrification rates of the 0% NaCl and 0.5% NaCl biofilm cultures, which increased by 72 and 34%, respectively, compared to the Ref300N-23C biofilm cultures. However, in the 1% NaCl biofilm cultures, these rates decreased by one third compared to the Ref300N-23C biofilm cultures (Table 3). The 5% NaCl and 8% NaCl biofilm cultures were strongly affected by poor biofilm development on the carriers, which was 50–100 less abundant in biomass than in the Ref300N-23C biofilm cultures (Table 3). In these biofilm cultures, NO₃⁻ took more than 24 h to be consumed and NO₂⁻ consumption was minimal. This result contradicts what was obtained in the first set of assays where the Ref300N-23C biofilm cultures subjected for 3 days to 5% and 8% NaCl did not affect the denitrification performance. To investigate this further, another Ref300N-23C biofilm cultures were derived, but after the fifth transfer cultures, the carriers were transferred in 5% NaCl ASW medium (here named 2.75–5% NaCl) and cultivated for another three carrier transfers (20 days). These biofilm cultures were able to sustain this high salt concentration. Compared to the Ref300N-23C biofilm cultures, the 2.75–5% NaCl biofilm cultures had a 74% increase in the denitrification rates. Because the biomass was twice the amount in these cultures, the specific denitrification rates were similar to those found in the Ref300N-23C biofilm cultures (Table 3).

The PCR-DGGE migrating profiles showed the persistence of strain NL23 in the 0% NaCl and 0.5% NaCl biofilm cultures. In the 1% NaCl biofilm cultures, however, strain NL23 was barely visible in DGGE profiles (Fig. S4). Persistence of strain NL23 was confirmed by qPCR where its concentrations in the 0% NaCl, 0.5% NaCl and 1.0% NaCl biofilm cultures (1.1 to 5.3 × 10⁴ napA cp/ng DNA) were at the same order of magnitude than in the OB (8.7 ×10⁴ napA cp/ng DNA) (Fig. 3B). As observed before, the concentrations of strain NL23 decreased by two to three orders of magnitude in the Ref300N-23C, 5% NaCl, 8% NaCl and 2.75–5% NaCl biofilm cultures (4.3 to 18 × 10¹ napA cp/ng DNA) compared to the OB (Fig. 3B). The levels of M. nitratireducenticrescens strains JAM1/GP59 increased by one order of magnitude in the 0% NaCl and 0.5% NaCl biofilm cultures (ca. 4 × 10⁴ narG1 cp/ng DNA) compared to the OB (5.6 × 10³ narG1 cp/ng DNA), at similar levels than strain NL23 (Fig. 3B). However, in the 1% NaCl biofilm cultures, the concentrations of M. nitratireducenticrescens strains JAM1/GP59 increased by another order of magnitude (1.5 × 10⁵ narG1 cp/ng DNA), which is similar to that found in the Ref300N-23C biofilm cultures (2.3 × 10⁵ narG1 cp/ng DNA) (Figs. 3A and 3B). As observed before, the concentrations of strain JAM1 decreased by approximately one order of magnitude in these cultures (6 to 23 × 10¹ tagH cp/ng DNA) when compared to the OB (2.5 × 10³tagH cp/ng DNA) (Fig. 3B). The concentrations of strain GP59 in the 0% NaCl biofilm cultures (4.6 × 10² nirK cp/ng DNA) were at similar level as in the OB (1.6 × 10² nirK cp/ng DNA). It increased by one order of magnitude in the 0.5% NaCl biofilm cultures (1.2 × 10³ nirK cp/ng DNA) and by two orders of magnitude in the 1.0% NaCl biofilm cultures (2.5 ×10⁵ nirK cp/ng DNA), similar to the levels reached in the Ref300N-23C biofilm cultures (2.9 × 10⁵ nirK cp/ng DNA) (Figs. 3A and 3B). All these results showed that strain GP59 was favored in the ASW medium with increasing concentrations of NaCl, which was the opposite for strain NL23 in these cultures.

Cultivation of the original biofilm in the IO medium

The important decrease of strain NL23 and the enrichment of strain GP59 in the Ref300N-23C biofilm cultures suggest that the switch from the continuous-operating mode that prevailed in the Biodome reactor to the batch mode in our biofilm cultures could have generated these specific changes in the H. nitrativorans and M. nitratireducenticrescens populations. Another factor that could have influenced these changes is the ASW medium that we used instead of the Instant Ocean (IO) from Aquatic systems Inc. used by the Biodome (Table S1). The dispersed cells of the OB were therefore cultivated in the IO medium, under the same conditions than the Ref300N-23C biofilm cultures (performed in parallel). The NO₃⁻ and NO₂⁻ dynamics in the Ref300N-23C biofilm cultures were similar to those reported in Fig. 1A. The denitrification rates were calculated at 1.89 mM-NO_x h⁻¹, and the specific denitrification rates corresponded to 0.0684 mM-NO_x h⁻¹ mg-protein⁻¹ (Table 3). In the IO biofilm cultures, the denitrification rates (0.76 mM-NO_x h⁻¹) were 2.5 lower than in the Ref300N-23C biofilm cultures (Table 3). However, because less biomass developed in the IO biofilm cultures (Table 3), its specific denitrification rates (0.0661 mM-NO_x h⁻¹ mg-protein⁻¹) were the same as in the Ref300N-23C biofilm cultures.

As before, PCR-DGGE migration profiles (Fig. S5) and qPCR assays showed the decrease in strain NL23 concentrations during the five carrier transfers in the Ref300N-23C biofilm cultures. Surprisingly, the DGGE band corresponding to strain NL23 was still present in the IO biofilm cultures (Fig. S5). This result was confirmed by qPCR assays where the concentrations of strain NL23 in the IO biofilm cultures (Fig. 3A) were at similar levels (7.0 × 10⁴ napA cp/ng DNA) as in the OB (8.7 × 10⁴ napA copies/ng DNA). In the IO biofilm cultures, the concentrations of strain JAM1 and strain GP59 were both approximately one order of magnitude lower than those of strain NL23 (Fig. 3A).

Cultivation of the original biofilm under oxic conditions and with NO${}_2^{-}$

The dispersed cells of the OB were cultivated under the same conditions of the Ref300N-23C biofilm cultures except that the cultures were incubated under oxic conditions. Growth was not impaired in these cultures, but the specific denitrification rates were 5 times lower than in the Ref300N-23C anoxic biofilm cultures (Table 3). Culturing the OB with a mix of 200 mg-NO₃⁻-N/L and 200 mg-NO₂⁻-N/L (here named 200-200N) generated biofilm cultures with denitrification rates 2.5 times lower than those in the Ref300N-23C biofilm cultures (Table 3). The proportions of M. nitratireducenticrescens strains JAM1/GP50 and strain NL23 in these two types of cultures were similar to the ones in the Ref300N-23C biofilm cultures (Figs. 3A and 3B).