Co-Occurrence and Cooperation between Comammox and Anammox Bacteria in a Full-Scale Attached Growth Municipal Wastewater Treatment Process

Cooperation between comammox and anammox bacteria for nitrogen removal has been recently reported in laboratory-scale systems, including synthetic community constructs; however, there are no reports of full-scale municipal wastewater treatment systems with such cooperation. Here, we report intrinsic and extant kinetics as well as genome-resolved community characterization of a full-scale integrated fixed film activated sludge (IFAS) system where comammox and anammox bacteria co-occur and appear to drive nitrogen loss. Intrinsic batch kinetic assays indicated that majority of the aerobic ammonia oxidation was driven by comammox bacteria (1.75 ± 0.08 mg-N/g TS-h) in the attached growth phase, with minimal contribution by ammonia-oxidizing bacteria. Interestingly, a portion of total inorganic nitrogen (∼8%) was consistently lost during these aerobic assays. Aerobic nitrite oxidation assays eliminated the possibility of denitrification as a cause of nitrogen loss, while anaerobic ammonia oxidation assays resulted in rates consistent with anammox stoichiometry. Full-scale experiments at different dissolved oxygen (DO = 2 – 6 mg/L) setpoints indicated persistent nitrogen loss that was partly sensitive to DO concentrations. Genome-resolved metagenomics confirmed the high abundance (relative abundance 6.53 ± 0.34%) of two Brocadia-like anammox populations, while comammox bacteria within the Ca. Nitrospira nitrosa cluster were lower in abundance (0.37 ± 0.03%) and Nitrosomonas-like ammonia oxidizers were even lower (0.12 ± 0.02%). Collectively, our study reports for the first time the co-occurrence and cooperation of comammox and anammox bacteria in a full-scale municipal wastewater treatment system.


. Preparation of Batch Assay Set-up and Experiments
Stock solutions were prepared as 10,000 mg NH 4 -N/L (using ammonium chloride) and 10,000 mg NO 2 -N/L (using sodium nitrite). Prior to conducting experiments, the IFAS media was collected from the aerobic zone and washed overnight in secondary clarifier effluent to remove loosely associated biomass. 1 L glass beakers were set up on stir plates with a magnetic bar placed inside along with an aeration stone connected to an aquarium pump to provide aeration for aerobic assays. After washing with secondary effluent, 40 pieces of IFAS media were added to the beakers with 900 mL of secondary clarifier effluent for the attached phase assays while 900 mL mixed liquor were added to beakers for suspended phase assays. The aquarium pump was turned on to provide aeration for aerobic experiments with pump settings modified based on DO measurements to achieve a DO concentration between 6-8 mg/L. Alkalinity and pH in all beakers were measured before the experiments and adjusted using the sodium bicarbonate stock solution to ensure adequate alkalinity for nitrification. Blank controls were established for each condition using 900 mL of filtered water.

SI-4
Section SI-2. Sample characterization and total solids measurements DO concentrations and temperature were measured directly at each sampling location in the fullscale IFAS system using a probe (IPM insiteIG with PortaCaddie) submerged in the wastewater. pH was measured using the HACH Company Pocket Pro+ pH meter (Cat. No. 9532000) immediately after sample collection. Duplicate samples of mixed liquor were filtered at all locations in the field through 0.45 M syringe filters for chemical analysis. Hach Company TNT Vials were used to determine concentrations of ammonia (TNT832, TNT831), nitrite (TNT839, TNT840), nitrate (TNT835), chemical oxygen demand (TNT821, TNT820), and total alkalinity (TNT870) of the filtered aqueous samples from both full-scale and intrinsic kinetic experiments. All samples were analyzed on a HACH DR3900 photospectrometer (Cat. No. LPV440.99.00002). IFAS media and suspended solids from both ends of the aerobic zone were collected on the sampling days to measure total solids (TS) and total suspended solids (TSS). Suspended solids and IFAS media were also collected from both ends of the aerobic zone for DNA extraction for microbial characterization.
Three IFAS pieces and suspended solids samples were taken from two aeration zones (beginning and end) on three different sampling dates. The suspended solids concentration (g/L) was measured using Standard Method 2540D. All IFAS pieces (n=18) were baked overnight at 105 C and then weighed before scrubbing off all biomass with a 2 N H 2 SO 4 solution and small bristle brush. The clean IFAS pieces were dried overnight at 105 C and then weighed. The mass of total solids (TS) on individual pieces were obtained by subtracting the mass of the cleaned IFAS piece from the mass recorded for dried attached biomass. The average mass of TS per media piece for each sampling date was then multiplied by 40 (number of IFAS pieces used in each assay) to estimate the total mass of attached biomass in batch assays for the attached phase. The average concentration of total attached biomass in grams per liter was estimated by dividing the average total mass of attached biomass (g) by 0.9 L (volume of secondary clarifier effluent used in batch assays). Total attached biomass in the aerobic zone of the treatment train was estimated by multiplying the average mass of attached biomass obtained from all 18 pieces by the number of IFAS pieces in the aerobic zone (~114155600). The average concentration of total attached biomass (g/L) was estimated by dividing the aforementioned average (g) by the tank volume, 984207 L.

Section SI-3. Sample collection and processing for full-scale experiments
The six sampling locations selected included the anoxic zone (n=3, one per stage for R1, R2, and R3), aerobic zone (n=2, one in the beginning (R4_Z1) and one at the end (R4_Z4)), and deaeration (n=1, R5) (Figure 1 main manuscript). Samples were collected using a long handle scooper approximately six hours after the DO concentration was modified in the aerobic zone, and only one DO setting was applied per day. In the field, IFAS media was stored by cutting pieces with a sterile razor blade and placed two per tube into 15 mL falcon tubes for DNA extraction. Both IFAS media and suspended solids samples for DNA extraction were placed in a -20 C freezer at the treatment plant until they were shipped on dry ice to Northeastern University, where they remained in a -80 C freezer until extraction. Samples were taken at five designated sampling locations along the transect on days when the DO was set to 2, 4, and 6 mg/L for measuring volatile fatty acids (VFAs). Seven VFAs, including acetic, butyric, caproic, isobutyric, isovaleric, propionic and valeric acid, were surveyed using standard method SM5560D, but only acetic acid was detected above the limit of quantification (LOQ = 5 mg/L) in any of the wastewater samples.

Figure SI-3:
Nitrogen species as a percent of influent total inorganic nitrogen (TIN) along the transect of the IFAS system. The facets are labelled based on the DO setpoints. Bars were generated by taking the average concentrations of ammonia (navy), nitrite (orange), and nitrate (green) between sample duplicates and dividing them by the average concentration of influent total inorganic nitrogen. This was done for each sampling location at all DO settings (2-6 mg/L). The average concentration of influent TIN was the summation of ammonia, nitrite, and nitrate concentrations acquired from sample duplicates at R1. Profiles of inorganic nitrogen indicated similar compositions of nitrogen species (mostly as ammonia) entering the aerobic zone when the DO was set to 2, 3, and 4 mg/L, but a higher proportion of nitrite and nitrate entered the zone when the DO was set in 6 mg/L. In this case, nitrate production in the aerobic zone was the highest out of all DO set points, and as a result, more nitrate was returned to the anoxic zone (R2/R3) through the internal recycle line. A dilution effect was observed between R1 and R2 on each sampling day due to the internal recycle coming into the anoxic zone at R2. For example, the average influent ammonia concentration (R1) on one sampling day was 19.45 mg NH 3 -N/L, and it was diluted to 13.7 mg NH 3 -N /L in R2.