Cometabolic removal of organic micropollutants by enriched nitrite-dependent anaerobic methane oxidizing cultures

Cometabolic removal of organic micropollutants by enriched nitrite-dependent anaerobic methane oxidizing cultures Miguel Martínez-Quintela, Adrián Arias, Teresa Alvarino, Sonia Suarez, Juan Manuel Garrido, Francisco Omil Accepted Manuscript How to cite: Journal of Hazardous Materials, 402 (2021), 123450. doi: 10.1016/j.jhazmat.2020.123450 Copyright information: © 2020 Elsevier B.V. This manuscript version is made available under the CC-BY-NC-ND 4.0 license (http://creativecommons.org/licenses/by-nc-nd/4.0) 1 Cometabolic removal of organic micropollutants by enriched nitrite- 1 dependent anaerobic methane oxidizing cultures 2 3 Miguel Martínez-Quintela1,*, Adrián Arias1, Teresa Alvariño1,2, Sonia Suárez1, Juan 4 Manuel Garrido1 and Francisco Omil1. 5 6 1Department of Chemical Engineering, School of Engineering, University of Santiago de 7 Compostela, Campus Vida, E-15782, Santiago de Compostela, Spain 8 2Galician Water Research Center Foundation (Cetaqua Galicia). Emprendia Building, 9 University of Santiago de Compostela, Campus Vida, E-15782, Santiago de Compostela, 10 Spain 11 12 *Corresponding author: Miguel Martínez-Quintela. E-mail: [email protected] 13 14 15 16 17 18 2 Abstract 19 The innovative and recently discovered n-damo process, based on anaerobic methane 20 oxidation with nitrite, was developed in a membrane-based bioreactor and evaluated in terms 21 of organic micropollutants (OMPs) removal. The main singularity of this study consisted in 22 the evaluation of organic micropollutants (OMPs) removal in the biological reactor. A 23 strategy consisting on progressively increasing the nitrogen loading rate in order to increase 24 the specific denitrification activity was followed to check if the selected OMPs were co- 25 metabolically biotransformed. Significant nitrite removal rate (24.1 mg N L-1 d-1) was 26 achieved after only 30 days of operation. A maximum specific removal of 186.3 mg N gVSS- 27 1 d-1 was obtained at the end of the operation, which is one of the highest previously reported. 28 A successfully n-damo bacteria enrichment was achieved, being Candidatus 29 Methylomirabilis the predominant bacteria during the whole operation attaining a maximum 30 relative abundance of about 40%. The natural hormones (E1 and E2) were completely 31 removed in the bioreactor. The specific removal rates of erythromycin (ERY), fluoxetine 32 (FLX), roxithromycin (ROX) and sulfamethoxazole (SMX) were successfully correlated 33 with the specific nitrite removal rates, suggesting a co-metabolic biotransformation. 34 35 Keywords: n-damo process, organic micropollutants, cometabolism, enrichment, MBR 36 37 38 39 40 41 3 42 1. INTRODUCTION 43 Anaerobic wastewater treatment processes have been largely applied in areas with temperate 44 and warm climates due to their lower energy consumption, generation of an enriched- 45 methane biogas and lower biomass production, relative to conventional aerobic systems. The 46 generated effluents can pose an environmental problem due to their high ammonia 47 concentration, and their significant content in dissolved methane, which is very often directly 48 desorbed into the atmosphere (Noyola et al., 2006; Souza et al., 2011). 49 Conventional biological nitrogen removal process has been applied to urban wastewater in 50 order to reduce water eutrophication in downstream water bodies. The main drawbacks of 51 this process (as requirement of a biodegradable carbon source, high energy consumption, 52 sludge production and greenhouse gas emissions) can be addressed with novel treatment 53 processes, such as those based on autotrophic anaerobic ammonium oxidation (anammox), 54 combined with nitrite/nitrate dependent anaerobic methane oxidation (n-damo), which have 55 proven to reduce effectively nitrogen from wastewater economically and with less 56 environmental impact (Liu et al., 2019). 57 The recently discovered n-damo bacteria (Candidatus Methylomirabilis Species: M. oxyfera; 58 M. sinica; M. lanthanidiphila), bacteria affiliated with the NC10 phylum, have demonstrated 59 their ability to perform both nitrogen and methane removal, by anaerobically oxidizing 60 methane using nitrite as electron acceptor (Ettwig et al., 2010; He et al., 2016; Versantvoort 61 et al., 2018). The slow growth of this microorganism, between 1 and 2 weeks, implies that 62 the enrichment process of this bacteria in a mixed culture may take several months (Allegue 63 4 et al., 2018; Ettwig et al., 2009). Thus, in order to avoid problems related with biomass 64 washout, n-damo bacteria enrichments have been successfully achieved using membrane 65 bioreactor-based configurations (MBR). Among the first results using this type of 66 configuration were those reported by Kampman et al. (2014) achieving a nitrite removal rate 67 of 36 mg N L-1 d-1. After further development, these results could be improved to nitrite 68 removal rates of 116 mg N L-1 d-1 by Allegue et al. (2018). 69 N-damo microorganisms (bacteria and archaea) are currently object of interest because of 70 their suitability to develop innovative processes in which methane and nitrogen compounds 71 can be simultaneous removed. Additionally, previous studies showed positive results at pilot 72 scale hybrid (anoxic-aerobic) reactors for the removal of organic micropollutants (OMPs) 73 (Arias et al., 2018). However, until now, the n-damo based technology is not suitable for its 74 implementation in WWTPs since it is not possible to achieve high denitrification activities 75 (Wang et al., 2017). Thus, more research about is needed to enhance their metabolic activity 76 as well as to analyze the possible effects on their development in the presence of other 77 substances like OMPs. 78 OMPs are currently considered as an important challenge to be addressed by the new 79 innovative technology developed, since they have been linked with several environmental 80 risks (Tran et al., 2018). The European Union started to advance in the regulation of these 81 contaminants in water (Sousa et al., 2018), by the establishment of a watchlist that identifies 82 8 compounds that should be monitored (last updated by Commission decision (EU) 2018/840 83 of 5 June 2018). It included in 2015 three macrolide antibiotics - erythromycin, azithromycin 84 and clarithromycin; three hormones – estrone, estradiol and ethinylestradiol; one anti- 85 inflammatory drug – diclofenac (DCF). Its update 2018 supposed among others the 86 5 incorporation of two additional antibiotics: amoxicillin (group of penicillins) and 87 ciprofloxacin (group of quinolones), as well as the removal of DCF. 88 Among the different factors which determine OMP removal in biological reactors, the redox 89 conditions have shown to highly influence their fate. Depending on the condition applied, 90 different microbial populations will grow, activating different metabolic routes and enzymes 91 and, consequently, determining the biotransformation route of the micropollutant. In general, 92 removal efficiencies are larger in aerobic conditions compared to anoxic ones, due to the 93 higher oxidation potential of the oxygen compared with NOx species (Alvarino et al., 2018). 94 The combination of different redox environments normally maximizes the final removal 95 efficiency (Arias et al., 2018). A lack of knowledge on the OMP biotransformation in anoxic 96 environments is still identified (Torresi et al., 2017). 97 OMPs are commonly present at trace levels in water bodies. Consequently, cometabolism is 98 considered as the main biotransformation mechanism, in which the presence of a growth 99 substrate is needed to promote the production of non-specific enzymes which can 100 accidentally biotransform OMPs (Fischer and Majewsky, 2014; Plosz et al., 2010; Tran et 101 al., 2013). One of the most studied cometabolic processes are linked to nitrification, in which 102 the ammonium monooxygenase enzyme has proven to have an active role on several OMP 103 biotransformation due to its low specificity (Fernandez-Fontaina et al., 2016). Under anoxic 104 environments, Polesel et al. (2017) found positive correlations between the specific 105 heterotrophic denitrifying activity and the biotransformation rate constants of different non- 106 recalcitrant compounds. Until now, few studies about cometabolism were performed under 107 anaerobic or anoxic environments (Gonzalez-Gil et al., 2019; Pomiès et al., 2015). As far as 108 6 the authors know, there are neither previous studies about OMP removal in processes based 109 on the use of n-damo microorganisms. 110 The aim of this study was to achieve an enriched n-damo bacteria culture in an MBR-based 111 configuration and to assess the removal efficiency of OMPs in the system. The research was 112 focused on monitoring the reactor performance at increasing nitrogen loading rates in order 113 to evaluate the robustness of the system in terms of conventional parameters and bacterial 114 dynamics. Besides, the cometabolic biotransformation of OMPs was assessed by studying 115 the correlations between the denitrification kinetics and the OMP removal rates. 116 2. MATERIALS AND METHODS 117 2.1 Reactor configuration and operation 118 A 10 L lab-scale MBR was used to carry out the enrichment of n-damo culture, with a 119 working volume of 6.65 L (Fig. S1). Due to low activities of these microorganisms, complete 120 biomass retention in the reactor was achieved by using a submerged hollow-fiber 121 ultrafiltration membrane module (Puron), which was operated in cycles of 7 min of 122 permeation and 0.5 min of relaxation. The surface of the membrane was 0.5 m2 and the pore 123 size 0.03 µm. A pressure sensor PN2069 (IFM) was used to record permeation and relaxation 124 transmembrane pressures and an Atlas Scientific pH Probe was placed inside the reactor. The 125 reactor was provided with a thermostatic bath to maintain the temperature at 28 ºC. 126 The inoculum used was taken from a denitrifying pre-anoxic MBR treating an UASB 127 effluent, in which dissolved methane was used as carbon source for denitrification (Silva- 128 Teira et al., 2017). This biomass was composed of a complex microbial population and 129 characterized by a low presence of n-damo bacteria. A total suspended solids concentration 130 7 of 0.5 g L-1 was used as inoculum. The system was fed with a medium containing 9 mg L-1 131 of NH4Cl, 100 mg L-1 of NaHCO3 and NaNO2 as nitrite source according to biomass activity 132 capacity and requirements for this study. The addition of ammonium nitrogen was decided 133 according to previous studies suggesting that there is a need of supplying an additional 134 nitrogen source to promote n-damo bacteria growth, as they are unable to extract it from their 135 denitrification pathway (Wang et al., 2019). Macronutrients and trace compounds used in the 136 enrichment were the proposed by Allegue et al. (2018), and the same feeding strategy was 137 performed. Aluminum foil feeding bags, which are totally impermeable to oxygen, were used 138 and feeding was deoxygenated to avoid the presence of oxygen by using nitrogen gas during 139 10 min. A programmable PLC Micro 820 (Allen-Bradley) connected to a computer was used 140 to control the operation of the system. Apart from the liquid feed, 9 L d-1 of a gas mixture, 141 composed of 95% of CH4 and 5% of CO2, were supplied to the system by using a mass-flow 142 controller 4800 Series (Brooks). CO2 provided buffer capacity to the system as well as the 143 sodium bicarbonate added in the liquid feed. A 5.5 L min-1 recirculation gas from the top of 144 the reactor to the bottom using a mini laboratory blower N 86 KT. 18 (Laboport) made 145 available methane and CO2 in excess in order to avoid electron donor or buffer limitation, 146 attain complete mixing and prevent membrane fouling. 147 Twelve OMPs known to be present in sewage, were spiked continuously to the synthetic 148 feed: three neurodrugs (fluoxetine FLX, carbamazepine CBZ, diazepam DZP), four 149 antibiotics (erythromycin ERY, roxithromycin ROX, sulfamethoxazole SMX, trimethoprim 150 TMP), three endocrine disruptors (estrone E1, β-estradiol E2, 17 α-ethinylestradiol EE2), and 151 two anti-inflammatories (ibuprofen IBP and naproxen NPX). They were spiked on day 35 at 152 a concentration of 1 ppb, to simulate environmental concentrations. Sampling was limited to 153 8 the liquid phase (feed and permeate), because of the low biomass concentration inside the 154 reactor and the limited affinity of the selected OMPs to sorb onto solids. 155 2.2 Analytical methods 156 Temperature, pH, nitrogen species, total suspended solids (TSS), volatile suspended solids 157 (VSS) and dissolved organic carbon (DOC) were measured according to Standard Methods 158 (Rice et al., 2012). The dissolved oxygen (DO) concentration in the feeding bags was 159 recorded by using a multi-parameter meter (Hach HQ40d) connected to a luminescent optical 160 probe (IntelliCAL LDO101). Gas flow was measured by using a Milli GasCounter MGC-1 161 V3.3 PMMA (Ritter) and its composition was measured in a gas chromatograph (HP 5890 162 Series II) with a Porapack Q 80/100 2 m × 1.8” (SUPELCO) column. 163 Microbial community was characterized by withdrawing a homogeneous biomass sample 164 from the reactor, fixing in 4% v/v paraformaldehyde for fluorescence in situ hybridization 165 (FISH) and keeping biomass samples frozen for DNA extraction. FISH was performed as 166 described by Regueiro et al. (2012), with specific probes for n-damo bacteria, damo archaea 167 and anammox bacteria. Sequencing of the 16s rRNA gene was carried out according to the 168 protocol described by Allegue et al. (2018). Raw sequences processing was carried out with 169 USEARCH v11.1. Clustering into Operational Taxonomic Units (OTUs) was performed with 170 the UPARSE algorithm taking into account a 97% of sequence similarity. 171 OMPs analysis samples were collected in amber glass bottles, prefiltered (AP3004705, 172 Millipore), stored at 4 ºC and pre-concentrated by using solid phase extraction (SPE) in 3 mL 173 OASIS HLB cartridges. Pre-concentrated samples were analyzed by LC-MS-MS for 174 antibiotics, hormones and neurodrugs, while anti-inflammatories were analyzed by GC-MS. 175 15 most common organisms found in the enrichments of wastewater treatment technologies 282 (Kuenen, 2008). Although Luesken et al. (2011) proposed that anammox bacteria have higher 283 affinity for NO2- than n-damo bacteria, its minor presence in the whole culture of the reactor 284 suggests a residual contribution to denitrification . 285 According to the FISH analyses, the operational results in terms of nitrogen removal and the 286 16s RNA gene sequence (Illumina), an enrichment of the culture in n-damo bacteria was 287 clearly achieved. The relative abundance values of the NC10 phylum were similar that those 288 reported in Allegue et al. (2018), in a very similar enrichment procedure and system. Due to 289 the low abundance of other denitrifying groups, like anammox or heterotrophic denitrifying 290 bacteria, it is assumed that n-damo bacteria is responsible for almost all the nitrogen removed 291 in the MBR. 292 All the changes in the bacterial community occurred when the MBR performance suffered 293 operational issues; there was no evidence that the presence of the OMPs affected the 294 microbial community. 295 3.3 OMPs results 296 The OMPs removal efficiency was studied in the system at the end of each operational period 297 (Section 3.1) to determine their removal efficiency for the different specific activities 298 achieved in the reactor (Fig. 3) and maintained stable at least for 10-15 days. Sorption was 299 not considered as a relevant removal mechanism in this study because the studied OMPs are 300 not lipophilic, except for FLX (whose particular behavior is discussed in section 3.4). 301 Additionally, the concentration of VSS during the MBR performance was very low (< 0.4 g 302 VSS·L-1) supporting the statement of low contribution of sorption to OMP removal. 303 16 According to the reported values of the Henry coefficients for the selected OMPs, neither 304 volatilization supposed a relevant removal mechanism (Suárez et al., 2008). Consequently, 305 the OMP removal efficiencies determined in this work were attributed to biotransformation. 306 Accordingly, the selected OMPs can be grouped in the following three categories: recalcitrant 307 compounds (removal efficiency < 20%) such as TMP, CBZ, DZP, EE2, IBP, NPX and DCF; 308 highly biotransformable substances (> 90%) including E1 and E2; and compounds 309 moderately biotransformed in the reactor (20-80%) as ERY, SMX, ROX and FLX. 310 311 Figure 3. Removal efficiencies of the selected OMPs in each stage. 312 As expected, the removal efficiency of CBZ and DZP barely reached 10% in any of the 4 313 sampling campaigns. Their recalcitrant behavior in the three redox environments was 314 previously pointed out in several studies (Alvarino et al., 2018; Suarez et al., 2010). 315 Regarding IBP, its biotransformation degree was around 15% in the last two stages. In this 316 work, IBP proved to be much more recalcitrant in anoxic environments compared to removal 317 efficiencies reported in aerobic reactors, with conventional heterotrophic bacteria 318 0% 20% 40% 60% 80% 100% ERY FLX ROX SMX TMP CBZ DZP E1 E2 EE2 IBP NPX Removal efficiency (%) Period 1 Period 2 Period 3 Period 4 17 (Fernandez-Fontaina et al., 2016). However, this low biotransformation degree in anoxic 319 conditions is in accordance with by previous batch experiments using a synthetic media with 320 acetate and nitrate (Kassotaki et al., 2018) or fed with pre-clarified wastewater with an 321 external nitrate addition (Torresi et al., 2017). Thus, we can confirm that IBP biodegradation 322 is highly influenced by redox conditions. According to Falås et al. (2013), the removal of this 323 compound is strongly influenced by the physical conformation of the biomass, achieving 324 better results with biomass attached to carriers than suspended in the mixed liquor. 325 The high removal of natural hormones as well as the recalcitrant behavior of the EE2, was 326 observed previously in a denitrifying activated sludge reactor (Suarez et al., 2010). 327 Nevertheless, the removal efficiencies obtained in this study for natural hormones are higher 328 (almost 100% of biodegradation versus 72%) than those reported by Suarez et al. (2010). No 329 effect was observed in the removal of natural and synthetic hormones during the reactor 330 operation. Despite having a similar chemical structure, a steric impediment was suggested in 331 other studies to explain the different removal efficiencies achieved for the hormones E2 and 332 EE2 (Czajka and Londry, 2006). 333 In the case of the moderately biodegradable compounds (ERY, ROX, SMX and FLX) the 334 removal efficiency observed during the reactor operation varied strongly (< 10% up to > 335 70%). This behavior will be further discussed in section 3.4. In the anoxic chamber of a pilot 336 plant which combined the three redox conditions, Arias et al., (2018) found a 30% removal 337 efficiency for ROX and ERY, comparable with the removal found in the aerobic 338 compartment. Moreover, Burke et al. (2014) observed a preferential biodegradation of ROX 339 under anoxic compared to aerobic conditions. In the case of SMX, its removal is commonly 340 associated with anaerobic conditions (Alvarino et al., 2018; Arias et al., 2018). Nevertheless, 341 18 significant biotransformation rates were reported in anoxic denitrifying conditions in batch 342 tests, in agreement with the results obtained in this research (Kassotaki et al., 2018; Polesel 343 et al., 2017; Torresi et al., 2017). According to Torresi et al. (2017), when the primary 344 substrate of the microbial consortia is available, the removal rate of SMX is enhanced, 345 suggesting that the main biotransformation mechanism of this compound is cometabolism. 346 Concerning FLX, some authors suggested that its removal is mainly attributed to sorption in 347 anoxic environments (Alvarino et al., 2018; Pomiès et al., 2015). This could explain the high 348 removal efficiency observed for FLX already in P1 (Fig. 3) as further discussed in the next 349 section. On the other hand, Suarez et al. (2010) reported a high removal of FLX in an anoxic 350 denitrifying reactor with a strong influence of the SRT on its biodegradation. Thus, the 351 removal of this compound could be enhanced with time in the reactor as no purges of biomass 352 were carried out throughout its operation. 353 3.4 Cometabolic biotransformation of antibiotics 354 The behavior of FLX, ROX, ERY and SMX could suggest removal by cometabolism, 355 according to the increasing trend in their removal efficiencies along the different operational 356 periods. Since the OMPs concentration fed to the bioreactor was the same during the whole 357 operation, if the biodegradation rate of one compound increased in parallel to the n-damo 358 primary metabolism we suggest that this pollutant was cometabolically biotranformed. In 359 order to analyze this, the specific removal rates of the primary substrate (nitrite) and the 360 OMPs were plotted in Fig 4. For this purpose, the data from the OMP concentrations in the 361 feed and in the permeate, the applied HRT and the measured VSS concentration in the reactor 362 were considered for each period of operation. The denitrifying specific activity in each stage 363 was specified in the section 3.1. 364 19 365 366 Line equation R2 ERY y: 0.0344x - 3.3274 0.9999 FLX y: 0.0152x - 1.1011 0.9598 ROX y: 0.0361x - 3.5831 0.9885 SMX y: 0.0133x - 0.5420 0.9686 367 Figure 4. (a) Specific micropollutant removal rate for ERY, FLX, ROX and SMX at 368 different specific biomass activity. (b) Linear regression equation and correlation 369 coefficient (R2) for each compound (y is the specific removal rate of each compound and x 370 is the specific removal rate for nitrite) and considering the data for periods P2-P4. 371 372 The linear correlations in Fig. 4b showed a clear relationship between the or specific primary 373 substrate removal rate and the specific removal rate of ERY, SMX, ROX and FLX between 374 P2 and P4. This implies that best results in terms of OMPs biotransformation kinetics were 375 achieved in P4, when the specific denitrification capacity of the n-damo bacteria was the 376 highest. Even in the case of SMX, for which a lower removal efficiency (45.6%) was 377 measured in P4 (Fig. 3), the specific removal rate in this period was higher than in the 378 0.0 1.0 2.0 3.0 4.0 ERY FLX ROX SMX rOMP (µg/VSS· d) P1 P2 P3 P4 a) b) 20 previous operational periods, supporting cometabolic features. Between the first and the 379 second sampling campaign, there were no significant differences in the specific removal rate 380 of the OMPs, with the exception of FLX. If cometabolism is the main biodegradation 381 mechanism, some authors suggest that enough growth substrates are necessary to stimulate 382 the creation of the relevant enzymes for the degradation of the OMPs (Tran et al., 2013). This 383 may have occurred in the first sampling period (P1), when the nitrite concentration fed to the 384 system was the lowest of the total reactor operation (20 mg N L-1 d-1 of NRL). So, for that 385 reason the results obtained in the period 1 were not included when considering the 386 cometabolic data analysis (Fig. 4). 387 In the case of FLX, the higher specific removal rate achieved in P1 compared to the other 388 OMPs may be due to its sorption onto the sludge in the MBR. Considering a Kd of 1000 L 389 kgTSS-1 (Alvarino et al., 2016) for this compound, the removal would be almost completely 390 associated to such mechanism (concentration in feed: 1 ppb, 0.51 gTSS L-1 in P1) since its 391 global removal efficiency was around 50% (Fig. 3). The sorption coefficient of this 392 compound may be slightly higher in anoxic environments, however the order of magnitude 393 is the same and sorption is more dependent on the physico-chemical charactheristics of the 394 micropollutants and the biomass conformation than to redox environment (Alvarino et al., 395 2018). Pomiès et al. (2015) exposed that FLX removal in anoxic conditions was mainly due 396 to sorption, with little influence of biodegradation. However, once sorption equilibrium is 397 established, normally after several days of contact between the sludge and the OMPs, 398 biotransformation becomes the main removal mechanism (Yang et al., 2011). Thus, from P1 399 onwards, the removal of FLX could be attributed fully to biodegradation, due to the 400 21 negligible biomass growth during the whole reactor operation and only values from P2 to P4 401 have been considering for the correlation with the specific nitrite removal of the reactor. 402 In previous studies, the removal of SMX has been associated with the activity of aerobic 403 autotrophic and heterotrophic bacteria. Its biotransformation rate was positively correlated 404 with the presence of organic matter as main substrate (Alvarino et al., 2018; Fernandez- 405 Fontaina et al., 2016). However, in our study no correlation has been found between the 406 presence of the primary substrate (nitrite) with the microcontaminants removal rate. Polesel 407 et al. (2017) reported a clear correlation between the specific denitrification rate and the 408 biotransformation constant of SMX and ERY (the correlation coefficient reported is greater 409 than 0.95) in batch assays with heterotrophic denitrification bacteria, in agreement with our 410 results. Their denitrification range varied between 10 and 50 mg N gTSS-1 d-1, consequently 411 focusing on lower levels than the present research (50-190 mg N gVSS-1 d-1). In Fig. 4b) 412 different slopes were determined for the OMP studied, showing that ERY and ROX were 413 more influenced by the cometabolic effect than SMX and FLX. This may indicate that 414 cometabolism is more dependent to the specific denitrification rates, as it was previously 415 reported (Polesel et al., 2017; Torresi et al., 2018), than to the presence of high concentration 416 of the primary substrate as in P2 and P4 the NRR was almost the same (around 40 mg N L-1 417 d-1) but the specific nitrite removal rate in such periods was different. 418 It is worth to note that most results on OMP removal under anoxic conditions have been 419 published for conventional denitrification processes. Since the nitrogen removal metabolism 420 of the n-damo process is different from conventional denitrification (Fig. 5c), the behavior 421 of OMP in those processes could be different. In conventional denitrification, the nitrate is 422 progressively reduced to nitrite, nitric oxide, nitrous oxide and, finally, to nitrogen gas. 423 22 Specially the nitrite reduction pathway is different, due to the presence of and NO dismutase 424 (NOD) which performs the last step of the denitrification and provides O2 for the methane 425 oxidation (Fig. 5a) (Wu et al., 2011). The common enzyme between both processes, the 426 copper nitrite reductase (EC 1.7.2.1) (NIR), was correlated with the biotransformation rate 427 constant of some compounds like SMX, ERY or TMP and the abundance of the gene 428 encoding for these enzyme (Torresi et al., 2018). However, there are other enzymes in the 429 methane oxidation pathway like monooxygenases (particulate methane monooxygenase 430 (pMMO) (EC 1.14.18.3)) or dehydrogenases (methanol dehydrogenase (MDH) (EC 1.1.2.7)) 431 which can be other candidates to biotransform OMPs, as they are known to participate in 432 some pollutant degradations such as toluene or xylene (Jindrová et al., 2002). Further 433 research needs to be done to elucidate which enzymes in the n-damo metabolism are 434 responsible for these biotransformations. 435 436 Figure 5. Candidatus M. oxyfera metabolic pathways. (a) Nitrite pathway, (b) Methane 437 pathway. (c) Conventional heterothrophic denitrification chain. Abbreviations: NIR; nitrite 438 reductase (EC 1.7.2.1); NOD, nitric oxide dismutase; pMMO, particulate methane mono- 439 oxygenase (EC 1.14.18.3); MDH, methanol dehydrogenase (EC 1.1.2.7) ; mtdB, 440 methylene-H4MPT dehydrogenase (EC 1.5.1.5); FDH, fromate dehydrogenase (EC 441 1.17.1.10); NAR, nitrite reductases (EC 1.7.1.1); NOR, nitric oxide reductase (EC 1.7.5.2); 442 NOS, nitrous oxide reductase (EC 1.7.2.4). 443 444 445 4. CONCLUSIONS 446 23 This research provided new knowledge about the capability of n-damo cultures to 447 biotransform OMPs. A successful n-damo enrichment was achieved along 412 d of operation 448 in a MBR. Limiting the quantity of oxygen entering the MBR, by sealing efficiently the 449 reactor and deoxygenating the feeding bags, was revealed as a crucial factor in order to 450 achieve a fast n-damo bacteria enrichment and denitrification activity (24 mg N L-1 d-1 in 451 around 30 days). The system also showed its robustness, being able to recover the previous 452 denitrifying activity (40 mg N L-1 d-1), after an inhibitory period. The 16S rRNA gene 453 sequencing results showed that Methylomirabiliaceae family dominated the microbial 454 community with a relative abundance of 40 % during stable reactor operation. The OMPs 455 could be classified according to the achieved removal efficiencies in the reactor as: 456 recalcitrant, completely biodegradable and moderately biodegradable. For the moderately 457 biodegradable OMPs (ERY, ROX, SMX and FLX), a positive correlation between the n- 458 damo bacteria denitrification kinetics and the OMP specific biotransformation rates has been 459 found. This suggests that the main mechanism driving such biotransformation was 460 cometabolism. 461 5. ACKNOWLEDGEMENTS 462 This research was carried out with the financial support received from Spanish Ministry of 463 Economy and Competitiveness through the project COMETT (CTQ2016-80847-R), co- 464 funded by FEDER. M. Martínez would also like to express his gratitude to the same Ministry 465 for awarding a research scholarship (BES-2017-080503). The authors belong to the Galician 466 Competitive Research Group GRC (ED431C 2017/29), programme co-funded by FEDER, 467 and to CRETUS Strategic Partnership (ED431E 2018/01). 468 24 469 References 470 Allegue, T., Arias, A., Fernandez-Gonzalez, N., Omil, F., Garrido, J.M., 2018. Enrichment 471 of nitrite-dependent anaerobic methane oxidizing bacteria in a membrane bioreactor. 472 Chem. Eng. J. 347, 721–730. https://doi.org/10.1016/j.cej.2018.04.134 473 Alvarino, T., Suárez, S., Garrido, M., Lema, J.M., Omil, F., 2016. A UASB reactor coupled 474 to a hybrid aerobic MBR as innovative plant configuration to enhance the removal of 475 organic micropollutants. Chemosphere 144, 452–458. 476 https://doi.org/10.1016/j.chemosphere.2015.09.016 477 Alvarino, T., Suarez, S., Katsou, E., Vazquez-Padin, J., Lema, J.M., Omil, F., 2015. 478 Removal of PPCPs from the sludge supernatant in a one stage nitritation/anammox 479 process. Water Res. 68, 701–709. https://doi.org/10.1016/j.watres.2014.10.055 480 Alvarino, T., Suarez, S., Lema, J., Omil, F., 2018. Understanding the sorption and 481 biotransformation of organic micropollutants in innovative biological wastewater 482 treatment technologies. Sci. Total Environ. 615, 297–306. 483 https://doi.org/10.1016/j.scitotenv.2017.09.278 484 Arias, A., Alvarino, T., Allegue, T., Suárez, S., Garrido, J.M., Omil, F., 2018. An 485 innovative wastewater treatment technology based on UASB and IFAS for cost- 486 efficient macro and micropollutant removal. J. Hazard. Mater. 359, 113–120. 487 https://doi.org/10.1016/j.jhazmat.2018.07.042 488 Bhattacharjee, A.S., Motlagh, A.M., Jetten, M.S.M., Goel, R., 2016. Methane dependent 489 denitrificationfrom ecosystem to laboratory-scale enrichment for engineering 490 applications. Water Res. 99, 244–252. https://doi.org/10.1016/j.watres.2016.04.070 491 31 biodegradation of sulfonamide antibiotics by activated sludge: Experimental 636 assessment using batch data obtained under aerobic conditions. Water Res. 45, 3389– 637 3397. https://doi.org/10.1016/j.watres.2011.03.052 638 639 640