Understanding the enzymatic and cometabolic biotransformation of organic micropollutants under aerobic heterotrophic conditions

INTERNATIONAL DOCTORAL SCHOOL OF THE USC David Manuel Kennes Veiga PhD Thesis Understanding the enzymatic and cometabolic biotransformation of organic micropollutants under aerobic heterotrophic conditions Santiago de Compostela, 2022 Doctoral Programme in Chemical and Environmental Engineering TESIS DE DOCTORADO Understanding the enzymatic and cometabolic biotransformation of organic micropollutants under aerobic heterotrophic conditions David Manuel Kennes Veiga ESCUELA DE DOCTORADO INTERNACIONAL DE LA UNIVERSIDAD DE SANTIAGO DE COMPOSTELA PROGRAMA DE DOCTORADO EN INGENIERÍA QUÍMICA Y AMBIENTAL SANTIAGO DE COMPOSTELA 2022 DECLARACIÓN DEL AUTOR/A DE LA TESIS D./Dña. David M. Kennes Veiga Título da tese: Understanding the enzymatic and cometabolic biotransformation of organic micropollutants under aerobic heterotrophic conditions Presento mi tesis, siguiendo el procedimiento adecuado al Reglamento y declaro que: 1) La tesis abarca los resultados de la elaboración de mi trabajo. 2) De ser el caso, en la tesis se hace referencia a las colaboraciones que tuvo este trabajo. 3) Confirmo que la tesis no incurre en ningún tipo de plagio de otros autores ni de trabajos presentados por mí para la obtención de otros títulos. 4) La tesis es la versión definitiva presentada para su defensa y coincide la versión impresa con la presentada en formato electrónico. Y me comprometo a presentar el Compromiso Documental de Supervisión en el caso que el original no esté depositado en la Escuela. En Santiago de Compostela, 21 de diciembre de 2021. Firma electrónica AUTORIZACIÓN DE LOS DIRECTORES / TUTORES DE LA TESIS Understanding the enzymatic and cometabolic biotransformation of organic micropollutants under aerobic heterotrophic conditions D. Juan M. Lema Rodicio, Catedrático de Ingeniería Química Dña. Marta Carballa Arcos, Profesora Contratada Doctora INFORMAN: Que la presente tesis, se corresponde con el trabajo realizado por D. David M. Kennes Veiga, bajo nuestra dirección/tutorización, y a utorizamos su presentación, considerando que reúne l os r equisitos exigidos en el R eglamento de Estudios de Doctorado de la USC, y que como directores de esta no incurre en las causas de abstención establecidas en la Ley 40/2015. De acuerdo con lo indicado en el Reglamento de Estudios de Doctorado, declaramos también que la presente tesis doctoral es idónea para ser defendida en base a la modalidad de Monográfica con reproducción de publicaciones, en los que la participación del doctorando fue decisiva para su elaboración y las publicaciones se ajustan al Plan de Investigación. En Santiago de Compostela, 21 de diciembre de 2021 AGRADECIMIENTOS La presente tesis se ha podido llevar a cabo gracias a los proyectos de investigación COMETT (CTQ2016-80847-R) y ANTARES (PID2019-110346RB-C21) y a la ayuda de apoyo a la etapa predoctoral de la Xunta de Galicia (ED481A-2018/113). En primer lugar, me gustaría agradecer a mis directores de tesis, Juan y Marta, por haberme dado la oportunidad de realizar este doctorado. Vuestros consejos, apoyo y confianza han sido fundamentales para desarrollar esta tesis y superar los contratiempos que se han ido presentando durante el camino. He aprendido y evolucionado mucho durante estos años, tanto a nivel personal como profesional, y en gran parte se debe a vuestra supervisión. Muchas gracias. I would also like to thank Kathrin Fenner for accepting me for a doctoral stay at the EAWAG, in the Department of Environmental Chemistry. I really enjoyed those 6 months, I learnt a lot and I could understand a little better the biotransformation mechanisms of the micropollutants. I am also very grateful for the moments I shared with the people of the Department, especially with Bernadette and Anastasia, who greatly helped me in the adaptation process to a new group and country. Me siento muy afortunado por haber podido compartir este doctorado con gente maravillosa del Biogroup, que me ayudó y acompañó en esta aventura. Gracias a Lorena por sus enseñanzas y colaboración durante el doctorado; a Sabela y Alba por adentrarme en el bonito mundo de la microbiología; y a Mar, Mónica y Rosa por facilitar enormemente el desarrollo de esta tesis. Quiero agradecer también a Cristina su excelente labor como mentora, ofreciéndome siempre sus consejos y apoyo. Además, ha sido un placer y una suerte poder compartir estos años con Miguel, Andrés, Paula, Chechu, Lucía, Matías, Alba, Ricky… Mis mejores y más alegres recuerdos han venido de vuestra mano en los innumerables cafés, charlas, comidas, viajes… Finalmente, muchas gracias a mis padres por haberme inculcado el amor por la ciencia y la investigación y por haberme animado y acompañado durante el doctorado; y a María, mi hermana, por ser la alegría de la casa y traer siempre los momentos más divertidos. Y, por supuesto, también a Sara, por sus incondicionales ánimos, paciencia, comprensión y cariño que han sido tan importantes para llegar hasta aquí. Sin todos vosotros, la culminación de esta tesis habría sido mucho más difícil, por lo que el éxito de este trabajo también es vuestro. ¡Muchas gracias! 2.3.4. Dissolved oxygen, temperature and pH .................................................................................. 55 2.3.5. Organic Micropollutants.......................................................................................................... 55 2.3.6. Transformation product analysis and structure elucidation .................................................... 57 2.3.7. DNA ........................................................................................................................................ 58 2.3.8. Proteomics............................................................................................................................... 59 2.3.8.1. Proteome extraction......................................................................................................... 59 2.3.8.2. Protein detection.............................................................................................................. 59 2.3.8.3. Protein data analysis........................................................................................................ 60 2.4. CALCULATIONS............................................................................................................60 2.4.1. Calculations of reactor performance ....................................................................................... 60 2.4.2. OMPs mass balance................................................................................................................. 61 CHAPTER 3. The organic loading rate affects organic micropollutants’ cometabolic biotransformation kinetics under heterotrophic conditions in activated sludge.....65 3.1. INTRODUCTION............................................................................................................66 3.2. MATERIALS AND METHODS .....................................................................................67 3.3. RESULTS AND DISCUSSION.......................................................................................68 3.3.1. Biotransformation yield........................................................................................................... 68 3.3.2. Specific biotransformation rate ............................................................................................... 70 3.3.3. Influence of heterotrophic activity on the OMPs biotransformation constant........................ 73 3.4. CONCLUSION ................................................................................................................80 CHAPTER 4. Heterotrophic enzymatic biotransformations of organic micropollutants in activated sludge .............................................................................81 4.1. INTRODUCTION............................................................................................................82 4.2. MATERIALS AND METHODS .....................................................................................82 4.3. RESULTS AND DISCUSSION.......................................................................................83 4.3.1. Reactor performance ............................................................................................................... 83 4.3.2. Heterotrophic OMPs biotransformation in activated sludge systems ..................................... 86 4.3.3. Identification of enzymatic biotransformations of OMPs in aerobic heterotrophic conditions ........................................................................................................................................................... 89 4.4. CONCLUSION ................................................................................................................96 CHAPTER 5. Sulfamethoxazole triggers specific enzymatic activities under aerobic heterotrophic conditions: A metaproteomic approach ..............................................97 5.1. INTRODUCTION............................................................................................................98 5.2. MATERIALS AND METHODS .....................................................................................98 5.3. RESULTS AND DISCUSSION.......................................................................................99 5.3.1. SMX biotransformation under aerobic heterotrophic conditions............................................ 99 5.3.2. TPs identification under aerobic heterotrophic conditions.................................................... 101 5.3.3. Impact of SMX on microbial community structure .............................................................. 102 5.3.4. Impact of SMX on the metaproteome................................................................................... 105 5.4. CONCLUSION ..............................................................................................................110 CHAPTER 6. GENERAL DISCUSSION AND CONCLUSIONS .........................111 6.1. MAIN OUTCOMES OF THE THESIS.........................................................................112 6.1.1. Are aerobic heterotrophs effective to biotransform OMPs? ................................................. 113 6.1.2. Transformation products as a tool to identify key enzymatic activities................................ 114 6.1.3. Omic techniques as a tool to identify key microbial and enzymatic players........................ 115 6.2. MAIN GAPS AND FUTURE PERSPECTIVES...........................................................117 REFERENCES ............................................................................................................121 LIST OF PUBLICATIONS........................................................................................145 1 LIST OF ACRONYMS ADBI: Celestolide AEBSF: 4-benzenesulfonyl fluoride hydrochloride AHTN: Tonalide AMO: Ammonia monooxygenase AOA: Ammonia oxidizing archaea AOB: Ammonia oxidizing bacteria ATU: Allylthiourea BES: 2-bromoethanesulfonate BPA: Bisphenol A CBZ: Carbamazepine COD: Chemical oxygen demand DCF: Diclofenac DZP: Diazepam E1: Estrone E2: 17E-estradiol EE2: 17D-ethinylestradiol ERY: Erythromycin FAS: Ferrous ammonium sulphate FLX: Fluoxetine HHCB: Galaxolide HRT: Hydraulic retention time IBP: Ibuprofen LOQ: Limit of quantification NED: N-(1-naphthyl)ethylenediamine NOB: Nitrite-oxidizing bacteria NP: 4-nonylphenol NPX: Naproxen OCT: Octyne OLR: Organic loading rate OMP: Organic micropollutant OP: 4-octylphenol ROX: Roxithromycin 2 SPE: Solid phase extraction SMX: Sulfamethoxazole SRT: Sludge retention time TCA: Tricarboxylic acid TCS: Triclosan TMP: Trimethoprim TP: Transformation product TSS: Total suspended solids USE: Ultrasonic solvent extraction VSS: Volatile suspended solids WWTP: Wastewater treatment plant RESUMO RESUMO Resumo 4 Capítulo 1. Introdución e contexto Os microcontaminantes orgánicos (MCOs) representan unha extensa gama de sustancias de consumo crecente en todo o mundo que inclúe produtos farmacéuticos, de uso persoal e hormonas, entre outros. Estes compostos, a pesar de estar presentes en concentracións moi baixas, no rango de ng a μg L-1, supoñen un risco tanto para o medioambiente como para a saúde humana a causa da súa persistencia, bioacumulación e actividade biolóxica. Ademais, poden acentuar fenómenos de toxicidade, mutaxenicidade, resistencia antibiótica e estres oxidativo, así como causar desordes dos sistemas nervioso, hormonal e reprodutivo. Os MCOs rematan nas estacións depuradoras de augas residuais (EDAR), que os reciben a través de diferentes rutas, como nas augas residuais urbanas, hospitalarias e industriais. Non obstante, a pesar de ser capaces de reducir as concentracións de moitos compostos, as EDAR non foron deseñadas para eliminar os MCOs, que frecuentemente son emitidos ao medioambiente sen tratar. Ademais, as aplicacións, propiedades físicoquímicas e persistencia dos MCOs é moi ampla e as condicións medioambientais e de proceso aplicadas (como temperatura, pH, comunidade microbiana ou potencial redox, entre outros) son fundamentais para reducir a súa concentración, o que da lugar a grandes diferenzas na súa presenza e eficiencia de eliminación nas EDAR. A eliminación dos MCOs nos procesos biolóxicos das EDAR ocorre principalmente por (i) volatilización, (ii) sorción ao lodo e (iii) biotransformación, aínda que para algúns compostos tamén se poden producir procesos de transformación abiótica. A etapa de biotransformación está considerada como o mecanismo principal xa que a volatilización e a sorción soamente son relevantes para compostos que presentan altos valores da constante da lei de Henry e do coeficiente de partición sólido-líquido, respectivamente. Dita biotransformación pode producirse metabólicamente se a enerxía xerada a partir da reacción de transformación dos MCOs é suficiente para promover o crecemento e mantemento celular. Sen embargo, a causa das baixas concentracións ás que están presentes os MCOs nas EDAR, a biotransformación xeralmente se produce cometabólicamente. Durante o cometabolismo, os microorganismos transforman compostos que non son de crecemento e non poden ser utilizados como única fonte de nutrientes e enerxía na presencia obrigada dun substrato primario. Este último atópase a maiores concentracións no medio, serve como doador de electróns e pode manter a Resumo 5 poboación microbiana e inducir a produción de enzimas e cofactores que son capaces de biotransformar os MCOs grazas a súa versátil actividade catalítica e a similitudes entre a estrutura química dos MCOs e a dos substratos naturais da enzima. Os procesos de biotransformación frecuentemente dan lugar á mineralización incompleta dos MCOs, que son convertidos en produtos de transformación (PT) que poden posuír os mesmos ou incluso maiores efectos adversos para o ecosistema que os seus compostos de orixe. Deste xeito, para reducir os riscos da descarga dos MCOs ao medioambiente e para maximizar a súa eliminación, é fundamental comprender os mecanismos de biotransformación que ocorren nos compartimentos biolóxicos das EDAR. Este estudo debe prestar especial atención aos PTs formados e buscar identificar os microorganismos e actividades enzimáticas que xogan un rol principal no proceso de biotransformación. O tratamento biolóxico máis común nas EDAR baséase nun sistema de lodos activados, que normalmente implica actividades heterótrofas e autótrofas nitrificantes. Mentres que a segunda foi amplamente estudada a causa do relevante rol xogado pola enzima amonio monooxixenasa, o papel dos microorganismos aerobios heterótrofos na biotransformación dos MCOs aínda non está claro a causa da complexidade e variedade das súas actividades metabólicas. O estudo dos mecanismos heterótrofos de biotransformación é importante pola relevancia desta poboación para eliminar a materia orgánica e polo seu crecente interese en novas concepcións de EDAR enerxéticamente máis eficientes, onde as etapas A, que traballan a altas taxas de carga orgánica e curtos tempos de retención de sólidos, operan basicamente en condicións heterótrofas. Polo tanto, esta tese busca obter una mellor imaxe dos mecanismos que dan lugar a biotransformación cometabólica dos MCOs baixo condicións aerobias heterótrofas, intentando responder as seguintes cuestións: xSon capaces os organismos heterótrofos de biotransformar amplamente MCOs con diversas propiedades físico-químicas? xProdúcese cometabólicamente dita biotransformación? xUnha maior actividade heterótrofa afecta a eficiencia de biotransformación dos MCOs? E a súa cinética ou constante cinética de biotransformación? Resumo 6 xCales son as principais reaccións de biotransformación que levan a cabo os organismos heterótrofos? xExiste algunha poboación microbiana heterótrofa que xogue un papel destacado durante a biotransformación dos MCOs? xCales son as actividades enzimáticas máis relevantes durante a biotransformación cometabólica dos MCOs en condicións aerobias heterótrofas? Capítulo 2. Materiais, métodos e equipos Neste capítulo descríbese a alimentación utilizada para o desenvolvemento dos experimentos levados a cabo nos distintos tipos de reactor (en continuo, secuenciais en batch e en batch). Ademais, preséntanse os MCOs estudados ao longo da tese e os métodos analíticos utilizados para a súa determinación e a dos PTs formados. Adicionalmente, proporcionase unha descrición detallada dos métodos analíticos empregados para monitorizar o estado dos experimentos (demanda química de oxíxeno (DQO), especies de nitróxeno, concentración de sólidos, oxíxeno disolto, pH e temperatura) e para obter información microbiolóxica (ADN e proteínas), que serviu para identificar os microorganismos e enzimas clave involucrados na biotransformación dos MCOs. Finalmente, preséntanse os cálculos empregados para analizar os datos obtidos nos distintos capítulos. Capítulo 3. A velocidade de carga orgánica afecta a cinética de biotransformación cometabólica baixo condicións heterótrofas en sistemas de lodos activos O rol dos microorganismos aerobios heterótrofos durante a biotransformación dos MCOs nos sistemas de lodos activos aínda non está claro. Neste capítulo, búscase dar resposta a súa capacidade para biotransformar cometabólicamente un grupo de 15 MCOs con diversas propiedades físico-químicas engadidos en concentracións entre 1 e 40 μg L- 1, así como coñecer a influenza da actividade heterótrofa na cinética de biotransformación cometabólica. Con este obxectivo, operouse un reactor continuo de tanque axitado (CSTR) a catro velocidades de carga orgánica (VCO) diferentes (0.2, 0.4, 0.6 e 0.8 g DQO L-1 d-1) a través do uso de concentracións crecentes de substrato primario (acetato) e mantendo constante Resumo 7 o tempo de retención hidráulico (24 h). Adicionalmente, para asegurar unicamente a presenza de actividade heterótrofa, engadiuse aliltiourea, un inhibidor nitrificante, e mantívose o tempo de retención celular ao redor de 5 días. Ademais, usando como inoculo os respectivos reactores CSTR, tamén se levaron a cabo ensaios en batch con concentracións iniciais de substrato primario (acetato) de 0.2, 0.4, 0.6 e 0.8 g DQO L-1. O obxectivo destes ensaios foi observar se variacións na actividade heterótrofa microbiana, determinada como a actividade específica máxima de DQO, afectaban a cinética de biotransformación dos MCOs e, particularmente, a constante de biotransformación cinética (kbiol). En primeiro lugar, os resultados dos CSTR amosaron que a eficacia de biotransformación foi superior ao 80 % para a maioría dos MCOs, coa excepción de carbamazepina e diazepam (inferior a 20 % a causa da súa natureza recalcitrante), trimetoprim (35 %) e eritromicina (70 %), e que os valores se mantiveron constantes ás 4 VCOs estudadas. Estes datos confirman as altas capacidades dos microorganismos heterótrofos para biotransformar os MCOs. Ademais, observouse que a biodegradación da fonte de carbono primaria e a biotransformación dos MCOs ocorría de modo simultáneo e que maiores VCOs deron lugar a maiores velocidades específicas de biotransformación, destacando ao cometabolismo como mecanismo principal de biotransformación baixo as condicións estudadas. Por outra banda, os ensaios en batch permitiron observar que a maior tasa de biotransformación específica determinada nos reactores CSTR a maiores actividades heterótrofas ocorre a través de variacións en kbiol. Este resultado permite afirmar que a kbiol dun determinado composto non depende soamente das súas propiedades físicoquímicas e das condicións ambientais, senon tamén da actividade específica do lodo causada polo substrato primario. A correlación positiva entre kbiol e a actividade específica máxima de DQO foi observada para todos os MCOs non recalcitrantes, con excepción de dúas fragancias. Non obstante, a influenza da actividade heterótrofa foi desigual entre os compostos estudados, indicando que a súa acción é específica para cada MCO. En conclusión, neste capítulo amósase que maiores VCOs conducen a maiores tasas de biotransformación cometabólica dos MCOs a través do aumento dos seus respectivos valores de kbiol. Non obstante, isto non implica necesariamente unha mellora no rendemento de biotransformación dos compostos sempre que se aporte un tempo de CHAPTER 1 CHAPTER 1. INTRODUCTION AND CONTEXT SUMMARY Biotransformation of trace-level organic micropollutants (OMPs) by complex microbial communities in wastewater treatment facilities is a key process for their detoxification and environmental impact reduction. Therefore, understanding the metabolic activities and mechanisms that contribute to their biotransformation is essential when developing approaches aiming to minimize their discharge. This chapter addresses the relevance of cometabolic biotransformation and discusses the main enzymatic activities currently known to take part in the aerobic removal of OMPs in wastewater treatment plants. Furthermore, the most common methodologies to decipher such enzymes are discussed. This chapter, under the consent of the publisher (Elsevier), was redrafted after: Kennes-Veiga, Gonzalez-Gil, L., Carballa, M., Lema, J. M. (2021). Enzymatic cometabolic biotransformation of organic micropollutants in wastewater treatment plants: A review. Bioresource Technology, 344B, 126291. ISSN: 0960-8524. Further information can be found in the “List of publications” (page 145). Chapter 1 16 1.1. ORGANIC MICROPOLLUTANTS IN WASTEWATER TREATMENT PLANTS Organic micropollutants (OMPs) cover a wide range of substances of worldwide increasing consumption, including pharmaceuticals, personal care products, hormones and industrial chemicals (Luo et al., 2014). Despite being present at very low concentrations, in the ng to Pg L-1 range, they pose the potential to harm the environment due to their persistence, bioaccumulation and biological activity (Margot et al., 2015). Besides, they can foster toxicity, mutagenicity, antibiotic resistance and oxidative stress, as well as causing disorders of the nervous, hormonal and reproductive system (Bilal et al., 2019; Tiwari et al., 2017). The applications, physicochemical properties and persistence of OMPs are very broad, causing their presence and removal efficiency to vary greatly in wastewater treatment plants (WWTPs), which, although they have proven to be capable of reducing their concentrations, were not designed to remove them (Margot et al., 2015; Petrie et al., 2014). The extent to which OMPs are removed also depends on environmental and process conditions, including temperature, pH, wastewater composition, microbial community and redox potential, among others (Alvarino et al., 2018a; Cao et al., 2020; Gulde et al., 2014; Kruglova et al., 2016; Rios-Miguel et al., 2021). OMPs removal in biological processes of conventional WWTPs occurs mostly due to (i) volatilization, (ii) sorption to the sludge and (iii) biotransformation (Pomiès et al., 2013; Tiwari et al., 2017), although abiotic transformation processes may also take place for some compounds (Yu et al., 2018). Biotransformation is considered the main responsible mechanism since volatilization and sorption are only relevant for compounds showing high values of the Henry’s law constant or solid-liquid partition coefficient, respectively (Nguyen et al., 2021; Su et al., 2021). OMPs biotransformation could occur metabolically if the energy generated from such reaction is sufficient to promote biomass growth and maintenance. Differently, cometabolic biotransformation is the ability of microorganisms to degrade non-growth substrates that cannot be used as the sole nutrient and energy source in the obligate presence of a primary substrate. The latter is available at higher concentrations, serves as an electron donor and can maintain the microbial population and induce the production of enzymes and cofactors which are capable of biotransforming the OMPs thanks to their versatile catalytic activity (Fernandez-Fontaina et al., 2014; Fischer and Majewsky, 2014; Krah et al., 2016). Among the scientific Introduction and context 17 community, it is widely assumed that the main biotransformation mechanism in real environmental conditions is cometabolism (Fischer and Majewsky, 2014; Lema and Suarez, 2017; Tran et al., 2013). Either metabolically or cometabolically, biotransformation processes can lead to incomplete mineralization of OMPs, thus being converted into transformation products (TPs) that may pose even increased adverse effects to the aquatic ecosystems and human health (Gulde et al., 2016). Consequently, a thorough analysis of the fate and transformation reactions of the OMPs in the biological compartments of WWTPs is of vital importance to properly assess the environmental impacts (Men et al., 2017). For such reason, in recent times research has evolved from (i) trying to develop and optimize analytical methods to determine OMPs concentrations in different matrixes into (ii) assessing their occurrence in streams and compartments and determining the removal efficiencies of varying technologies and operational conditions and finally into (iii) attempting to comprehend the underlying removal mechanisms, metabolic pathways and enzymatic activities occurring in WWTPs during sorption and biotransformation processes (Alvarino et al., 2018a), which would greatly help to mitigate OMPs in WWTPs. 1.2. COMETABOLIC BIOTRANSFORMATION OF OMPS IN WWTPs Previous studies have shown that many OMPs can be biotransformed metabolically or cometabolically (Fig. 1.1) depending on their physicochemical properties and the environmental conditions (Tran et al., 2013). Making a distinction between metabolism and cometabolism is important to have a better understating of the mechanisms involving OMPs biotransformation and to optimize their overall removal. However, multiple studies do not discriminate between them because differentiation can be difficult, particularly when working with mixed microbial communities such as activated sludge. Indeed, metabolism and cometabolism may coexist (Mandarić et al., 2018); sometimes is unclear if energy derives from OMPs biodegradation or dead cells (Tran et al., 2013); and metabolic and cometabolic steps can be interrelated and substitutable since they are part of a metabolic network evolved as a whole by the microbial community (Fischer and Majewsky, 2014). In this sense, microorganisms are continuously developing new catabolic pathways for substrates to access energy and nutrients or to detoxify compounds, and can develop new metabolic tools which may allow to turn cometabolic Chapter 1 18 processes into metabolic ones (Fischer and Majewsky, 2014; Kolvenbach et al., 2014). In fact, promiscuous enzymes, which can catalyze reactions with multiple substrates, have been associated with both processes that occur intentionally (metabolism) and fortuitously (cometabolism) (Fischer and Majewsky, 2014). The metabolism of OMPs is challenging because growth-linked degradation reactions can only proceed when reaction thermodynamics and kinetics are favorable. To date, only heterotrophic microorganisms have proven to be capable of carrying out such process and, although the OMPs concentration threshold required for metabolic activities remains unclear, it appears to be at levels considerably higher than those typically detected in environmental samples and WWTPs (Nsenga Kumwimba and Meng, 2019; Tran et al., 2013). Thus, OMPs biodegradation may not be energetically favorable to microorganisms, requiring promiscuous enzymes to biotransform them thanks to structural similarities with primary substrates, even if such reactions tend to proceed at much lower rates (Kolvenbach et al., 2014; Lema and Suarez, 2017; Nguyen et al., 2021). As a consequence, in WWTPs, it is assumed that OMPs are biotransformed cometabolically as a side effect, linked to the elimination of macropollutants such as organic carbon, nitrogen and phosphorous (Fischer and Majewsky, 2014) and, perhaps, to endogenous decay when growth substrates are depleted (M. H. Kim et al., 2020). During cometabolism, the initial growth to non-growth substrate concentration ratio is important since high values may help microorganisms to recover from toxicity caused by OMPs and TPs and because initial substrate concentrations determine whether sufficient reducing power could be diverted to non-growth substrate biotransformation and cell growth and maintenance (M. H. Kim et al., 2020; Rios-Miguel et al., 2021). Accordingly, higher cometabolic biotransformation efficiencies could be expected with increased primary substrate loading rates, as observed in nitrifying and heterotrophic activated sludge (Fernandez-Fontaina et al., 2012; Helbling et al., 2012; Kennes-Veiga et al., 2020). However, it is not always the case since there is an upper (saturation limit) and lower (activation limit) threshold of primary substrate loading rates influencing cometabolism (Carneiro et al., 2020; Gonzalez-Gil et al., 2018b; Sheng et al., 2021). Loading rates below that threshold lead to insufficient reducing power supply for cometabolism and values above the threshold reach a biotransformation limit due to a cometabolic decoupling between OMPs and primary substrates (Sheng et al., 2021) or because of thermodynamic constraints such as chemical equilibrium or enzymatic reversibility Introduction and context 19 (Gonzalez-Gil et al., 2019a, 2018a, 2018b). Interestingly, some studies have also observed that under starvation conditions or low availability of easily degradable carbon, biotransformation may improve thanks to the stimulation of multiple microorganisms and an increase in the abundance or diversity of some enzymes (Achermann et al., 2018b; Nguyen et al., 2021). The specificity constant ratio, or kinetic efficiency, of non-growth to growth substrates of a microbial culture, which characterizes the inherent competition, is also key during cometabolism since competition for the enzymatic active site may occur, leading to hindered biotransformation of some OMPs due to the higher concentrations at which growth substrates are present (Gonzalez-Gil et al., 2021; M. H. Kim et al., 2020; Plósz et al., 2010). Non-competitive inhibition may also occur when OMPs and primary substrates are not structurally analogous and bind to different active sites, resulting in reduced overall oxidation rates (Lema and Suarez, 2017; Su et al., 2015). Furthermore, the presence of specific moieties in OMPs may facilitate or hinder cometabolism and influence initial biotransformation reactions that condition the likelihood of subsequent reaction steps and the overall biotransformation rate and efficiency (Helbling et al., 2012; Kolvenbach et al., 2014; Nguyen et al., 2021; Tran et al., 2013). OMPs cometabolism has been observed in multiple environments. For instance, under nitrifying conditions, it was observed the cometabolic biotransformation of several OMPs along with ammonium oxidation (Gonzalez-Gil et al., 2021; Han et al., 2019; Sheng et al., 2021; Wang et al., 2019). Ammonia oxidizing bacteria (AOB) and archaea (AOA), the novel commamox population (Han et al., 2019) and slow-growing heterotrophs (Gonzalez-Gil et al., 2021) have been reported to participate in cometabolism, while other populations, such as nitrite-oxidizing bacteria (NOB), have shown a less relevant contribution (Yu et al., 2018). There is also evidence of cometabolism under heterotrophic conditions (Fernandez-Fontaina et al., 2016; Fischer and Majewsky, 2014; Gonzalez-Gil et al., 2021; Kennes-Veiga et al., 2020), although the carbon source of the primary substrate has a considerable influence on the biotransformation rate and extent (Larcher and Yargeau, 2011; Oliveira et al., 2019; Torresi et al., 2016a). The combination of nitrifying and heterotrophic activities has been shown to improve OMPs biotransformation and to reduce TPs accumulation thanks to the higher microbial diversity and a broader expression of non-specific enzymes (Fernandez- Fontaina et al., 2016; Khunjar et al., 2011). In contrast, information about cometabolism Chapter 1 under anoxic conditions is scarce despite evidence of their equal or even higher capabilities to biotransform some OMPs (Falås et al., 2013). Nonetheless, Torresi et al. (2017) proved cometabolism for several pharmaceuticals in denitrifying reactors and Martínez-Quintela et al. (2021), under the recently discovered N-Damo process, which consists of anaerobic methane oxidation with nitrite, also observed cometabolic biotransformation of multiple compounds. Similarly, anaerobic systems have also proven their capacity to reduce OMPs environmental impact in wastewater effluents (Arias et al., 2018; Phan et al., 2018) and it is believed that they could enhance the removal of some compounds poorly biotransformed under aerobic conditions (Harb et al., 2019; Lin et al., 2020). Gonzalez-Gil et al. (2017a, 2018b) and Carneiro et al. (2020) proved cometabolism for multiple compounds during anaerobic digestion; Alvarino et al. (2014) observed a linear relationship between the biotransformation rate of sulfamethoxazole, trimethoprim and naproxen and the methanogenic activity in an upflow anaerobic sludge blanket reactor and Oliveira et al. (2019) determined sulfamethazine cometabolism with anaerobic sludge using several carbon sources. Finally, OMPs cometabolism has also been observed in other biological processes, as in microalgae-based systems (Liu et al., 2021; Vo et al., 2020), in feast-famine conditions (Tang et al., 2021) and along with polyhydroxyalkanoates metabolism in a bioreactor set up for phosphorous removal (Torresi et al., 2019). 20 Introduction and context 21 Fig. 1.1. Representation of the metabolic and cometabolic biotransformation of OMPs. OMPs Waste products (CO 2 , H 2 O, TPs ) Energy (mobility, heat) Macromolecules (Cell components, growth, storage) Precursor molecules Nutrients Anabolism (synthesis reactions) ADP Energy ATP Catabolism (degradative reactions) Primary substrate Waste products (CO 2 , H 2 O) Energy (mobility, heat) Macromolecules (Cell components, growth, storage) Precursor molecules Nutrients ADP Energy ATP Catabolism (degradative reactions) OMPs TPs OMPs TPs Anabolism (synthesis reactions) ns ) Metabolism Cometabolism Chapter 1 22 1.3. ENZYMATIC ACTIVITIES INVOLVED IN THE BIOTRANSFORMATION OF OMPs UNDER AEROBIC CONDITIONS The metabolic capabilities and enzymatic network of the microbial communities present in WWTPs determine their biotransformation potential (Fischer and Majewsky, 2014; Helbling et al., 2012). However, the complexity of biological systems and the limited knowledge about enzymatic mechanisms has resulted in scarce information about the role of microbial strains and enzymes involved in biotransformation, which is crucial to develop risk-assessment tools that help predict OMPs pathways and half-lives in WWTPs (Achermann et al., 2020). Cometabolic biotransformation consists of a broad sequence of individual reactions and several OMPs may undergo various removal strategies concurrently, leading to a huge array of candidate enzymatic activities that may participate in biotransformation (Helbling et al., 2010a; Kolvenbach et al., 2014). Such enzymes are frequently believed to be located intracellularly, carrying out most biotransformation reactions thanks to their high metabolic versatility. However, extracellular enzymes are particularly required for the break-down of large and ionized OMPs that pose difficulties for cellular uptake and limit biotransformation rates; and they may be highly relevant to reduce the selection pressure for antibiotic-resistant genes (Krah et al., 2016; Zumstein and Helbling, 2019). Similarly, it is generally assumed that biotransformation occurs thanks to catabolic enzymes, but anabolic enzymes also seem to play a key role, highlighting the challenge of deciphering them (Achermann et al., 2018a; Stadler et al., 2018). The most common biological treatment in WWTPs is based on an activated sludge system where autotrophic nitrifiers and heterotrophs represent the most important microbial populations, although their relative contribution to OMPs biotransformation remains unclear (Lema and Suarez, 2017; Nguyen et al., 2021; Polesel et al., 2017). Nitrifying microorganisms can be classified into: AOB and AOA, involved in the oxidation of ammonia to nitrite, and NOB, capable of oxidizing nitrite to nitrate. Numerous studies performed in recent years have shown the key role played by nitrifying enzymes during biotransformation (Nsenga Kumwimba and Meng, 2019; Su et al., 2021) and, although further studies are needed to elucidate the role of promising enzymes such as hydroxylamine and nitrite oxidoreductases, enzymes present in AOB and AOA seem to greatly contribute to the process, while those present in NOB may play a minor role (Helbling et al., 2012; Su et al., 2021; Yu et al., 2018). Ammonia monooxygenase (AMO; Introduction and context 23 EC 1.14.99.39), present in AOB and AOA, is considered responsible for the biotransformation of multiple compounds (Fernandez-Fontaina et al., 2016; Men et al., 2017; Su et al., 2021; Wang et al., 2019; Yu et al., 2018), frequently through oxygen insertions resulting in hydroxylation reactions and sometimes in dehydrogenation or reductive dehalogenation (Helbling et al., 2012; Su et al., 2021). However, some studies have shown that the contribution to biotransformation of nitrifiers and AMO in activated sludge systems may often be over-estimated and belittle the role of other heterotrophic and nitrifying enzymes (Helbling et al., 2012; Men et al., 2017). In fact, the biotransformation of certain compounds in nitrifying activated sludge reactors could happen thanks to slow-growing heterotrophs (Achermann et al., 2018b; Gonzalez-Gil et al., 2021), suggesting that attention should also be paid to side sludge microbial activities. In this sense, high sludge retention times (SRT) are considered to improve OMPs biotransformation thanks to a broader microbial and functional diversity and a larger network of enzymatic activities produced by the additional presence of slow-growing microorganisms (Achermann et al., 2018b; Yuxin Wang et al., 2020). Achermann et al. (2018a) observed a characteristic trend between the SRT and the biotransformation of OMPs undergoing oxidation reactions, suggesting that those reactions may be catalyzed by enzymes less generally widespread among bacteria, that may become more abundant with the development of some groups of microorganisms. Differently, they observed a weak dependence between the SRT and the biotransformation of OMPs undergoing substitution reactions, indicating that enzymes broadly expressed by different microorganisms may be carrying them out, possibly because they are involved in mechanisms of central metabolism or general defense. Similarly, Torresi et al. (2018) did not find a positive correlation between some OMPs biotransformation and biodiversity, implying that the SRT and microbial richness may be only relevant when the biotransformation steps are performed by a limited number of taxa or enzymes. The capability of heterotrophic cultures to biotransform OMPs is very relevant and it has been proven that just their basal expression may be sufficient to achieve high removal extents of multiple xenobiotics (Fischer and Majewsky, 2014). Their biotransformation potential comes from their ability to express a huge number of different enzymatic activities thanks to the wide range of carbon sources that they can assimilate, making it a difficult task to determine their abundance and importance (Kennes-Veiga et al., 2021b). Numerous biotransformation reactions documented in aerobic experiments Chapter 1 1.4. APPROACHES TO DETERMINE KEY ENZYMES A huge amount of enzymes participate in biological wastewater treatments and, thanks to functional redundancy, the same metabolic process can be conducted by several microorganisms and enzymatic routes (Sambamoorthy and Raman, 2018). Therefore, linking the biotransformation of OMPs with specific enzymatic activities is a complex task that entails a deep understanding of the biological processes and the development of appropriate experimental strategies. In this section, the main approaches employed to identify the key enzymes are described and analyzed from a critical point of view (Fig. 1.2). 30 Introduction and context 31 Fig. 1.2. Frequent approaches applied to determine the enzymatic activities involved in OMPs biotransformation. Chapter 1 32 1.4.1. In vitro assays In vitro assays refer to a methodology where cell-free enzymes are cultured in a controlled and artificial environment outside of the living microorganism to address their role on OMPs biotransformation. So far, two approaches of this technique have been followed: (i) use of a sole purified enzyme, either obtained commercially (Gonzalez-Gil et al., 2017) or in the laboratory from a specific microorganism (Prior et al., 2010; Xu et al., 2015); and (ii) use of an enzymatic lysate directly extracted from the reactor biomass and composed of a mixture of enzymes (Gonzalez-Gil et al., 2019a; Krah et al., 2016; Zumstein and Helbling, 2019). The procedure for both strategies is summarized in Fig. 1.3. When using sole enzymes, significant efforts are needed to select the suitable ones. Firstly, the enzyme should be representative of the biological system under study and show a relevant activity in the bioreactor (Gonzalez-Gil et al., 2017), which requires a deep understanding of the metabolic pathways and species involved in the process. Secondly, it has to be feasible to obtain the enzyme in a purified form. In the case of using a cell-free lysate with a cocktail of enzymes directly extracted from the biomass reactor, different procedures can be conducted. If the goal is to study extracellular enzymes, cell lysis is not needed (Zumstein and Helbling, 2019), but it should be performed to extract intracellular enzymes (Krah et al., 2016). To recover membrane enzymes and disassemble enzymes from the extracellular polymeric substance, surfactants and cation exchange resins should be added, respectively. Once the cell-free lysate is obtained, the supply of cofactors and inhibitors should be considered to promote and suppress specific enzymatic activities, which could aid in the identification of specific enzymatic activities related to OMPs removal (Gonzalez-Gil et al., 2019a). The use of sole purified enzymes in in vitro assays allows to test and prove specific biotransformation pathways without the interference of other enzymes, which could happen when using an enzymatic lysate, allowing to narrow down the prediction and identification of TPs. Successful results with commercially purified enzymes were reported by Gonzalez-Gil et al. (2019b, 2017), who demonstrated that acetate kinase and hexokinase could transform OMPs with carboxyl and hydroxyl groups and moderate steric hindrance. According to Bisswanger (2011), thousands of enzymes exist, but less than 15 % are described in detail and only hundreds are commercially available. Even if so, enzymatic substrate selectivity and affinity can differ depending on the microorganism synthesizing the enzyme. For instance, acetate kinase from E. coli is more Introduction and context 33 substrate-specific than that from M. thermophila (Bock et al., 1999), which might result in a lower affinity for certain cosubstrates (i.e., OMPs) and explain why a reduced BPA biotransformation was achieved when using the former instead of the latter, which is present in anaerobic digesters (Gonzalez-Gil et al., 2019b, 2017). An alternative to commercially purified enzymes is to isolate microbial species from a bioreactor, identify the enzymes of interest and purify them to perform in vitro experiments. There are some studies where cytochrome P450 enzymes of specific bacteria were expressed in E. coli to perform in vitro experiments that demonstrated their ability to hydroxylate diclofenac (Prior et al., 2010; Xu et al., 2015); although the microorganisms studied do not belong to wastewater biological systems. This methodology enables testing a wider spectrum of enzymes, but its experimental success depends on complex challenges ranging from isolating and cultivating pure microorganisms to expressing and purifying the enzymes. In fact, the statement of Fischer and Majewsky (2014a) pointing out that the isolation of target enzymes present in WWTPs has not been documented, is still valid nowadays. The closest approach reported is the obtainment of a cocktail of extracted enzymes from biomass of activated sludge systems (Krah et al., 2016; Zumstein and Helbling, 2019) and anaerobic digesters (Gonzalez-Gil et al., 2019b), which allowed working with a wide variety of native enzymes. This method, despite being a simplification of in vivo processes, seems a more realistic strategy than working with individual enzymes. Yet, it might be difficult to ascertain which are the specific enzymes responsible for OMPs biotransformation. In fact, although the mentioned studies provided sound arguments about the candidate enzymes responsible for the observed biotransformation reactions, their identity could not be irrefutably proved. Moreover, another limitation of this procedure is that it is not possible to extract all the enzymes nor perform long-term experiments since the enzymes can rapidly lose their activity, which could compromise the comparison with results in vivo. Furthermore, in vitro assays are limited in properly representing complex biological processes, and it cannot be fully assured that reactions will occur in the same way and extent as under real bioreactor conditions, where biomass and substrate heterogeneity is huge and competition and inhibition mechanisms might happen among enzymes and microorganisms. Nonetheless, their results are undoubtedly valuable to solve the puzzle of OMPs biotransformation. Chapter 1 34 Fig. 1.3. Main stages needed to perform in vitro assays with sole purified enzymes or a cell-free lysate. 1.4.2. Use of inhibitors in microbial and enzymatic cultures The use of enzymatic inhibitors is a fast and simple approach to determine the relative contribution of specific enzymatic activities present in a microbial population. The methodology consists in linking a decrease in OMPs biotransformation to the activity of the inhibited enzyme, which offers an interesting starting point to unravel OMPs biotransformation pathways. Results can imply that the inhibited enzymes and microorganisms are responsible for the biotransformation (direct involvement) or that their inhibition leads to the subsequent suppression of the microbial and enzymatic activities carrying out the process (indirect involvement) (Men et al., 2017). This approach is particularly useful when the inhibitor is very enzyme-specific, making its selection a key aspect that requires knowledge on the inhibited enzymes and the associated biotransformation reactions. In fact, a common problem occurs when the enzymatic specificity of the inhibitor is broader than expected, possibly because it had not been tested before with as many enzymes as those present in wastewater mixed microbial cultures (Gonzalez-Gil et al., 2019b; Helbling et al., 2012). Thus, performing multiple experiments with an extended range of inhibitors that theoretically hinder the same reaction could help to increase confidence in the results (Men et al., 2017). The role of AOB in OMPs biotransformation has typically been determined through inhibition studies with allylthiourea (ATU) (Fernandez-Fontaina et al., 2016; Margot et al., 2016; Wu et al., 2020). However, Men et al. (2017), who used two ammonia oxidation Biomass Wash (buffer) , concentrate, homogenize Cell lysis (Ultrasonication or bead beating) Remove cellular remains (Centrifugation, filtration) Cell-free lysate In vitro assays with OMPs Addition of cofactors, inhibitors Addition of surfactants, cation exchange resins Study of the biological process (Species, functions, metabolic pathways) Select key enzymes Measure the enzymatic activity in the bioreactor Obtain purified enzyme (commercially, laboratory) Find optimal assay conditions In vitro assays with OMPs Sole purified enzymes Cocktail of extracted enzymes Measure protein content and enzymatic activities Introduction and context 35 inhibitors, ATU and octyne (OCT), concluded that ATU could overestimate the role of AOB by not being its inhibitory effect as specific as previously thought. They observed that 17 compounds showed significantly reduced biotransformation with ATU compared to OCT, particularly those containing thioether and phenylurea groups. Since they determined that ATU barely affected heterotrophic respiration and the essential heterotrophic enzymes implicated in energy conservation and central metabolism, they suggest that ATU may have inhibited other non-essential heterotrophic enzymes possibly involved in OMPs biotransformation. Inhibition studies have also been performed to elucidate the role of other enzymes and microorganisms. Krah et al. (2016) carried out experiments with extracted native enzymes from activated sludge and, using peptidase inhibitors, determined the involvement of peptidases in the hydrolysis of amide bonds present in OMPs. They used three inhibitors targeting the action of endopeptidases (E-64 for cysteine peptidases (EC 3.4.22.-); 4-benzenesulfonyl fluoride hydrochloride (AEBSF) for serine peptidases (EC 3.4.21.-) and pepstatin A for aspartic peptidases (EC 3.4.23.-)) and documented their involvement in the biotransformation of atenolol and bezafibrate; although they point out the need of performing further studies for confirmation since the inhibitors could have affected the activity of other non-endopeptidase enzymes. Gonzalez-Gil et al. (2019b) performed inhibition assays in vitro with extracted native enzymes from anaerobic sludge using E-64, AEBSF and pepstatin A as peptidase inhibitors and castanospermine and 2- bromoethanesulfonate (BES) as glycosidase and methanogenic inhibitors, respectively. They observed that the biotransformation of clarithromycin, climbazole, citalopram and erythromycin was impaired under the action of the peptidase inhibitors and that atenolol biotransformation was fully inhibited under the action of AEBSF, suggesting the participation of peptidase enzymes in their removal. However, they also reported that the biotransformation of acetyl-sulfamethoxazole was considerably inhibited under the action of BES even though methyl-CoM reductase is unlikely involved in the process, indicating that the action of the inhibitor is more unspecific than expected. Moreover, reduced removal of erythromycin and clarithromycin in assays with castanospermine was not observed despite being good candidates for glycosylase action due to their cladinose moiety, suggesting that other glycosylases not affected by the inhibitor could participate in the biotransformation (Gonzalez-Gil et al., 2019b). Chapter 1 36 1.4.3. Transformation product analysis As previously mentioned, biotransformation of most OMPs in WWTPs does not lead to complete mineralization, giving rise to the formation of TPs. Since OMPs biotransformation is mainly influenced by the affinity of their chemical structure with the unspecific enzymes expressed during primary substrate biodegradation (Han Tran et al., 2017), TPs identification and prediction would be possible if the enzymatic cycles were deeply understood and the structures of growth substrates and OMPs could be correlated. In like manner, TPs structure analysis and elucidation can be an excellent tool to obtain information on OMPs biodegradability and link enzymatic activities taking part during biotransformation (Table 1.1). However, until recently, scarce information was available due to the challenging task of identifying and quantifying unknown TPs present at extremely low concentrations in complex matrices (L. Zhang et al., 2013). Fortunately, the development and optimization of versatile analytical tools have helped to search and identify many suspect and non-target compounds with reasonable accuracy and certainty (Fenner et al., 2021; Gulde et al., 2016). Determining key enzymes through TPs analysis provides several advantages, such as: the experimental design is generally simple; lab results can realistically be extrapolated to WWTPs (Kern et al., 2010); the risks of finding and acquiring commercial enzymes for in vitro assays or of suffering enzymatic activity loss when working with enzymatic lysates are avoided (section 1.4.1.); and the substantial costs of other techniques are eluded (section 1.4.4.). For such reason, to date, most candidate enzymes involved in OMPs biotransformation have been suggested applying this methodology (Achermann et al., 2018b; Gulde et al., 2016; Kennes-Veiga et al., 2021b; Yu et al., 2018). However, TPs elucidation cannot lead to a direct and complete confirmation of the responsible enzymatic activities and certain constraints might affect TPs detection and identification through mass spectrometry approaches, such as when: they are present at concentrations below the detection limit; they possess molecular masses outside the mass range of the full scan; they are highly unstable and readily transformable; their structures are too simple and lead to analytical errors; they have compound-specific properties limiting the analytical ionization efficiency; they have structural isomers; interfering ions with the same mass as the TPs are present; and there is a lack of reference standards that allow achieving higher confidence levels in the proposed TPs structures (Helbling et al., 2010b; Kern et al., 2010; Schymanski et al., 2014). Thus, it is advisable to combine results Introduction and context 37 from several approaches, such as in vitro assays, enzymatic inhibition and TPs analysis, which allows determining key enzymes with much higher confidence, as performed to identify the contribution of serine proteases (EC 3.4.21.-) during the biotransformation of atenolol in anaerobic conditions (Gonzalez-Gil et al., 2019b). Filling the knowledge gap of TPs formation and linking them to responsible enzymes just with laboratory studies is a considerable challenge since such assays can be expensive and time-consuming, biotransformation is influenced by multiple process and environmental factors and microorganisms may develop new catabolic pathways over time (Kolvenbach et al., 2014; Men et al., 2017; Yuxin Wang et al., 2020). Therefore, the use of modeling tools, such as PathPred (Moriya et al., 2010), CRAFT (CRAFT, 2009), OECD Toolbox (OECD, 2020), enviPath (Wicker et al., 2016) and the EAWAG-PPS (EAWAG-BBD/PPS), which are based on biotransformation rules extracted from microbial metabolic pathways and enzymatic reactions reported in the literature, may provide an appropriate and cost-effective approach to help predict transformation reactions at the different molecular functional groups (Helbling et al., 2010b). They can contribute to building pathway knowledge, determine the biodegradability of OMPs and TPs, and perform risk assessments. Besides, they are extremely useful when creating lists of expected TPs in suspect-screening approaches (Achermann et al., 2018b; Gulde et al., 2016; Kennes-Veiga et al., 2021b). However, their main constraint is that they are exceptionally sensitive but poorly selective, leading to the prediction of too many TPs and causing numerous false positives, likely due to the limited data available of microbial processes and for not considering the effects caused by moieties surrounding the target functional groups in the biotransformation rules. Another limitation is that most rules and microbial pathways are based on literature information obtained from studies performed with pure or enriched cultures, unspecific environmental conditions or where xenobiotics are used as primary substrates, resulting in biotransformation reactions unlikely to happen during the cometabolic biotransformation of trace-level OMPs by mixed cultures (Gulde et al., 2016). Thus, combining modeling tools and lab experiments under defined and environmentally realistic conditions is necessary to improve the prediction of TPs and learn about new biotransformation pathways. 1.4.4. Omics approach The recent evolution and improvement of sequencing techniques, the increased knowledge on sequence information and the development of protein and genome Chapter 1 38 databases, such as Eawag-BBD/PPS and enviPath (EAWAG-BBD/PPS; Wicker et al., 2016), have allowed using meta-omics association studies to elucidate OMPs biotransformation pathways and set up hypotheses of potentially involved enzymatic activities (Table 1.1) (Krah et al., 2016). With that purpose, metagenomics, reporting which microorganisms are present in a sample and its metabolic potential; metatranscriptomics, offering precise information on the microbial functions happening at a given time; and proteomics analyses, showing the active enzyme pool present in the microbiome, have been increasingly performed. Metagenomics techniques have been widely applied in experiments with OMPs through modern high-throughput DNA sequencing technologies, while the use of metatranscriptomics and metaproteomics is more recent and there is debate about their suitability (Fenner et al., 2021). Metatranscriptomics has a higher sensitivity to detect low-abundance gene transcripts than metaproteomics to detect expressed proteins, and transcripts have been shown to correlate properly with the protein abundance level, but it generally implies a higher cost. Additionally, metaproteomics may provide a higher potential to obtain mechanistic insights thanks to describing the enzyme pool more directly (Achermann et al., 2020; Fenner et al., 2021), but protein identification requires metagenomic data from the samples, along with detailed interpretation and annotation (Fenner et al., 2021). The introduction and accumulation of OMPs in bioreactors affect microbial community structure and leads to changes in the expression levels of biodegradation genes and gene products, as observed by Harb et al. (2016) using high-throughput 16S rRNA gene sequencing, metatranscriptomics and gene databases. Thus, -omics association studies have been broadly used to find biomarkers that can help to understand and predict the capacities of microbial communities and the influence of specific WWTPs parameters (Helbling et al., 2012; David R. Johnson et al., 2015a). For example, Cydzil- Kwiatkowska et al. (2020) determined a linear correlation between BPA concentration and the expression levels of the gen bisdA, which encodes for ferredoxin. Similarly, Zhou et al. (2015), in an experiment with Sphingobium sp. BiD32 at varying concentrations of BPA, studied its pathway and key enzymes through genomic, proteomic and TPs analysis. They observed the upregulation of 43 proteins belonging to the dehydrogenase, dioxygenase, hydratase, hydroxylase and cycloisomerase class; confirmed that phydroxybenzoate hydroxylase was involved in the first biotransformation step and identified the respective TP as a genetic biomarker for BPA biotransformation. Helbling Introduction and context 39 et al. (2015), with a multivariate model and bacterial 16S rRNA analysis, determined that specific phylogenetic groups can also serve as biomarkers of microbial community activity towards OMPs biotransformation, although taxonomic biomarkers may only be useful when there is no functional redundancy and biotransformation is not extensively distributed among taxa. In this sense, Vuono et al. (2016), analyzing 16S rRNA genes and rRNA gene expression, observed that rare taxa had higher ratios of rRNA to rDNA and a superior protein synthesis potential, which suggests their relevant role in reactor performance and OMPs biotransformation, and confirms that functional and taxonomic richness can positively influence OMPs removal (Stadler et al., 2018). The main advantage of meta-omics techniques is that they intend to characterize the complete set of genes or gene products present in a microbial community at a given time, avoiding the need for the specific and robust knowledge required for the previously described approaches (sections 1.4.1, 1.4.2 and 1.4.3) (Achermann et al., 2020; Krah et al., 2016). Besides, meta-omics association not only informs about key enzymatic activities but also points towards specific microorganisms responsible for biotransformation and provides information on the subcellular location and organismal origin of candidate enzymes (Krah et al., 2016). Such knowledge is highly important since it can allow predicting changes in OMPs biotransformation rates and pathways (Stadler et al., 2018) given that: (i) biotransformation steps may be carried out by different organisms (Achermann et al., 2020; Helbling et al., 2015; David R. Johnson et al., 2015a); (ii) enzymes of different microbial origins can possess varying catalytic activities (Gonzalez-Gil et al., 2019b, 2017; David R. Johnson et al., 2015a); (iii) the same TP of an OMP can be produced by different microorganisms and enzymes (Zhou et al., 2015); and (iv) the same gene may associate with different enzymatic activities depending on the reactor configuration (Harb et al., 2016). The essential limitation of -omic association mining is the generation of a large number of false positives that lead to non-causal correlations, adding great complexity to the identification of genes or gene products of interest if hypotheses are not reduced when the large datasets are obtained (Achermann et al., 2020; David R. Johnson et al., 2015b). Moreover, a direct statement on the involvement of an enzyme in OMPs biotransformation is often not possible since, for instance, the upregulation of transcripts does not always imply an increase in enzymatic activity and the physiological functions of promiscuous enzymes cannot be consistently predicted based on the knowledge of the Chapter 2 46 bioreactor are also present in WWTPs that deal with more complex carbon substrates. Moreover, other trace nutrients were also added to promote the growth of aerobic heterotrophic microorganisms (Table 2.3) and allylthiourea solution was added in the feeding with a concentration of 5 mg L-1 to avoid nitrification. Table 0.2. Feeding composition and concentrations used to obtain a chemical oxygen demand (COD) of 0.2 g L-1. In the experiments with other COD concentrations, the concentrations were calculated proportionally. Compounds mg L-1 Sodium acetate (CH 3 COONa·3H 2 O) 210 Acetic Acid (CH 3 COOH) 95 Ammonium chloride (NH 4 Cl) 40 Potassium dihydrogen phosphate (KH 2 PO 4 ) 10 Calcium chloride (CaCl 2 ) 5.5 Magnesium sulfate (MgSO 4 ) 5.5 Table 0.3. Trace metals composition and concentrations used in the feeding of the aerobic heterotrophic reactor. Compounds mg L-1 FeCl 3 ·6H 2 O 0.15 H 3 BO 3 0.015 CuSO 4 ·5H 2 O 0.003 KI 0.003 ZnSO 4 ·7H 2 O 0.012 CoCl 2 ·6H 2 O 0.015 MnCl 2 ·4H 2 O 0.012 2.2. EXPERIMENTAL DESIGN Different types of reactor configurations (continuous, sequential batch and batch reactors) were operated to understand the cometabolic biotransformation of a range of OMPs under aerobic heterotrophic conditions. Materials, methods and equipment 47 2.2.1. Continuous reactors 5 L continuously stirred lab-scale reactors connected to a 2 L settler were used in Chapters 3 and 4, while a reactor of 2 L connected to a 1 L settler was used in Chapter 5 (Fig. 2.1.). The reactors were inoculated with sludge from a conventional activated sludge reactor of a WWTP near Santiago de Compostela (Spain), setting an in-reactor biomass concentration of approximately 1.0 g volatile suspended solids (VSS) L-1, and operated at 25ºC. The WWTP is designed for 184000 population equivalents, receives an influent COD ranging between 0.2 - 0.7 g L-1 and operates with approximate SRT and hydraulic retention time (HRT) values of 10 d and 8 h, respectively. In the reactors, the HRT was set to 1 d to provide sufficient time to the microorganisms to achieve the maximum biotransformation they are capable of, at the same time as minimizing possible changes in the microbial population caused by an HRT too long. Besides, the SRT was maintained around 5 d to minimize the presence of slow-growing microorganisms and favor the heterotrophic activity (Achermann et al., 2018b). Fig. 2.1. Schematic representation of the continuous aerobic heterotrophic reactors. Aeration was provided to the reactors ensuring oxygen concentrations between 3.5 and 7.5 mg O2 L-1. After a start-up period of a few days, the OMPs were spiked in the feeding at the desired concentrations, which were selected according to typical WWTPs influent concentrations (Table 2.1.). The operation of the reactors was mainly monitored through the standard analytical methods (APHA, 2017) described in section 2.3. Soluble COD, ammonium, nitrate and nitrite concentrations, total suspended solids (TSS), VSS, pH and temperature were measured two to three times per week in the reactor influent (feedstock) and effluent. Measurements of OMPs and TPs were also performed when steady state conditions were reached at the different experimental conditions applied. Additionally, in Chapter 5, samples were taken for genomic and metatranscriptomic analysis. Chapter 2 48 2.2.2. Batch experiments Batch assays were set to observe if variations in the heterotrophic microbial activity, determined as the maximum COD specific activity, affected OMPs biotransformation kinetics and, particularly, the kbiol (Chapter 3). The experiments were performed with COD initial concentrations of 0.2, 0.4, 0.6 and 0.8 g L-1 by varying the sodium acetate and acetic acid concentration and maintaining the micro and macronutrients used in the continuous reactor. The biomass was taken from the continuous reactors working at the respective organic loading rates (OLR) (Chapter 3) to ensure that the microbial population was adapted to the experimental conditions of the assays. For each batch, 18 flasks were prepared, three for each time point to have triplicates (0, 1, 3, 8, 24 and 48 h). At each of the mentioned times, 3 flasks were taken to determine the concentration of the OMPs. The VSS concentration was set in all cases to approximately 0.80 g VSS L-1 and the temperature and stirring were fixed at 25qC and 150 rpm (Innova 4300 Incubator Shaker –New Brunswick Scientific). Neutral pH was ensured using punctual additions of NaOH or HCl when necessary and an oxygen concentration above 4.5 mg O2L-1 was maintained during experimentation. To monitor the operation, the same analyses as in the continuous reactor (soluble COD, ammonium, nitrate and nitrite concentrations, TSS, VSS, pH and temperature) were conducted for each time point, except for the solids and nitrogen concentrations which were measured at times 0, 24 and 48 h. OMPs analysis was performed (solid and liquid phase) by taking samples from the flasks at each time point. 2.2.3. Sequential batch reactors In Chapter 5, aiming to determine the key microbial and enzymatic players involved in SMX biotransformation, a total of 18 sequential batch reactors with a working volume of 100 mL were operated in an incubator at neutral pH, 25ºC, 150 rpm and with oxygen concentrations ranging between 3-5 mg O2L-1 for 25 days. Six different SMX concentrations were tested in triplicates: 0 (control), 50, 250, 500, 1000 and 2000 μg L-1. The reactors were inoculated with activated sludge from the same WWTP as the continuous reactors and the operation was performed under sterile conditions. The content of the sequential batch reactors was centrifuged daily at 6000 rpm and 10ºC for 10 min to separate the biomass from the supernatant. Then, the exhausted supernatant was removed and new feed was added together with the spike of SMX Materials, methods and equipment 49 corresponding to each reactor. Finally, the flasks were placed again in the incubator to resume operation. All the process was performed using a fume hood to ensure sterile conditions. Ammonium, nitrate and nitrite concentrations, TSS and VSS were determined once per week, while the pH and acetate concentration were determined daily. Acetate was determined through gas chromatography using a DB-Wax-Agilent Technologies column (30mx0.250mmx0.25μm) while the other parameters were determined according to the standard analytical methods (APHA, 2017) described in section 2.3. Besides, samples from the feeding and the exhausted supernatant of the reactors were taken on days 2, 17 and 25 for SMX and TPs analysis. Finally, samples for proteomic (days 1 and 25) and 16S rRNA gene sequencing (day 25) analyses were also taken. 2.3. ANALYTICAL METHODS Several analytical methods were applied to the experimental samples, generally according to Standard Methods (APHA, 2017). 2.3.1. Soluble chemical oxygen demand The chemical oxygen demand (COD) is used to determine the concentration of the organic matter present in a liquid sample and is defined as the amount of oxygen required to oxidize the organic content. The COD was determined following a closed reflux titrimetric method based on the Standard Methods 5220C (APHA, 2017), but with minor modifications. Overall, the procedure uses a strong chemical oxidant (potassium dichromate) in an acid medium, along with a catalyst (silver sulphate) that improves the oxidation of some organic compounds. After digestion, the unreduced potassium dichromate is titrated with ferrous ammonium sulphate to determine its consumption. Then, using this value, the amount of oxidized organic matter present in the sample is calculated in terms of oxygen equivalents. The reaction is shown below: ଶ଻ଶି ൅൅ାଷା൅ଶ൅ଶ(1) The required reagents, which allow for a detection range between 0.09 and 0.9 g COD L-1, are listed below: xStandard potassium dichromate digestion solution: 10.216 g of K2Cr2O7 and 33 g of HgSO4are dissolved in 0.5 L of distilled water prior to the addition of Chapter 2 50 167 mL of concentrated H2SO4. Then, the solution is cooled down to room temperature and the solution is diluted to reach a total volume of 1 L. xSulphuric acid reagent (catalytic solution): 10.7 g Ag2SO4are added to 1 L of concentrated H2SO4. xFerroin indicator solution: 1.485 g of C18H18N2·H2O and 0.695 g of SO4Fe·7H2O are dissolved in 0.1 L of distilled water. xPotassium dichromate solution 0.05N: 1.226 g of K2Cr2O7are dissolved in 0.5 L of distilled water. xFerrous ammonium sulphate titrant (FAS) 0.035N: 13.72 g of Fe(NH4)2(SO4)2·6H2O are dissolved in distilled water before adding 20 mL of concentrated H2SO4and dissolving to a total volume of 1 L. Firstly, samples are filtered through a pore size filter of 0.45 μm (MF-Millipore, Millipore) to determine the soluble COD. Then, 2.5 mL of each sample are placed in 10 mL digestion glass vessels, to which 1.5 mL of the digestion solution and 3.5 mL of the sulphuric acid reagent are added slowly and carefully to avoid mixing of the reagents at this stage. Similarly, blanks are prepared using 2.5 mL of distilled water. All vessels are sealed with Teflon, tightly capped and mixed before placing them in the block digester (ECO 16, VELP Scientific) previously set at 150 ºC for two hours. After digestion, the tubes are cooled to room temperature and then each sample is transferred to a 50 mL Erlenmeyer flask which is magnetically stirred. Then, 1-2 drops of the ferroin indicator are added and the solution is titrated using a microburet with FAS until a change in color from light blue to red/orange is observed. To determine the normality of the FAS solution, 5 mL of the potassium dichromate standard solution are mixed with 5 mL of distilled water and 3.5 mL of the catalytic solution. Then, after cooling down the sample to room temperature, titration is performed as previously explained. Finally, the COD concentration is calculated taking into account the two equations below: ୊୅ୗ ൌ଴Ǥ଴ଶହ ୚ూఽ౏ (2) ቀ୥ ୐ቁൌሺ୅ି୆ሻ൉୒ూఽ౏൉଼ ୚൉ (3) Materials, methods and equipment 51 Where NFAS is the normality of the FAS solution (mol L-1); VFAS is the volume of the FAS solution consumed in the titration (mL); A and B are the volumes of the FAS solution consumed by the blank and the sample, respectively; V is the sample volume used before dilution; DF is the dilution factor applied to the sample and 8 stands for the conversion factor between COD and FAS (g COD mol-1 FAS). 2.3.2. Nitrogen compounds All nitrogen analyses were performed after filtering the samples (0.45 μm, MF- Millipore, Millipore) to remove solid particles, according to the Standard Methods (APHA, 2017). 2.3.2.1. Nitrite Nitrite concentration was determined spectrophotometrically using a method based on the production of a violet azo dye at a pH ranging between 2 and 2.5. It consists of a coupling between diazotized sulphanilamide and N-(1-naphthyl)ethylenediamine dihydrochloride (NED dihydrochloride), as described in the Standard Methods (APHA, 2017) under the 4500-NO2-B procedure (Colorimetric method). The required reagents are: xSulphanilamide solution: 10 g of sulphanilamide (C6H8N2O2S) are dissolved in 0.1 L of concentrated HCl and 0.6 L of distilled water. When the solution is cool, additional distilled water is added until reaching a total volume of 1 L. xNED solution: 0.5 g of NED (C10H7NH(CH2)2NH2·2HCl) are dissolved in 0.5 L of distilled water. The determination procedure consists of the addition of 0.1 mL of each reagent to 5 mL of sample. Then, at least 20 minutes of reaction time are required before measuring the colored sample at a wavelength of 543 nm in the spectrophotometer (Shimadzu UV- 1800). The quantification is done with a calibration curve, which is prepared by making dilutions in the range of 0-0.25 mg NO2--N L-1 from a stock solution of 10 mg NO2-L-1. Then, the absorbance of each sample is measured spectrophotometrically and related to the nitrite concentration (NO2--N L-1) (Fig. 2.2.). Chapter 2 52 Fig. 2.2. Calibration curve for the determination of the concentration of nitrite. 2.3.2.2. Nitrate Nitrate concentration was determined through measurements of UV absorption at the wavelengths of 220 and 275 nm, which allows avoiding errors caused by the absorbance of dissolved organic matter. The procedure corresponds to the 4500-NO3-B method listed in the Standard Methods (APHA, 2017). The required reagents are the following: xHCl 1 N: 8.3 mL of HCl (37% w/v, 1.19 g L-1) dissolved in 0.1 L of distilled water. xSulfamic acid: Commercial reagent (H3NSO3). The determination procedure consists of adding 100 μL of HCl 1 N and a pinch of sulfamic acid, which allows avoiding interferences in the measurement caused by nitrite, to 5 mL of sample. Then, the absorbance is measured at 220 and 275 nm using the spectrophotometer (Shimadzu UV-1800) and the nitrate concentration (NO3--N L-1) is calculated using the calibration curve (Fig. 2.3.). The absorbance related to nitrate is obtained according to Eq. 4. ሺଷିെሻൌଶଶ଴ െʹ൉ଶ଻ହ (4) αͲǤʹͺ͵ͶͶǦ ͲǤͲͲͳͻͲ ;αͲǤͻͻͻ 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 mg NO 2 --N/L Absorbance Materials, methods and equipment 53 Fig. 2.3. Calibration curve for the determination of the concentration of nitrate. The calibration curve is prepared using a 10 mg NO3- L-1 stock solution, which is diluted to obtain samples ranging between 0-3 mg NO3--N L-1. Then, the absorbance is measured spectrophotometrically and related to the nitrate concentration. 2.3.2.3. Ammonium A spectrophotometric methodology was selected to determine the concentration of nitrogen in the ammonium form. The analysis is based on the production of indophenol blue through the reaction of ammonia with salicylate and hypochlorite, under the presence of sodium nitroprusside (Bower and Holm-Hansen, 1980). Thus, the required reagents are the following: x Reagent A: 0.28 g of sodium nitroprusside (Na2(Fe(CN)5NO)) and 440 g of sodium salicylate (C7H5NaO3) dissolved in 1 L of distilled water. x Reagent B: 18.5 g of NaOH and 120 g of sodium citrate (Na3C6H5O7) dissolved in 1 L of distilled water. x Reagent C: 5% (w/w) commercial solution of sodium hypochlorite (NaClO). x Reagent D: Solution consisting of a mixture of reagent B and Reagent C in a 7:1 ratio. This solution remains stable for 1 h after preparation. The determination procedure consists of adding 600 μL of reagent A and 1 mL of reagent D to 5 mL of the previously filtered sample. Then, the mixture is left to react for 2 h in an environment protected from light. The measurement of the colored sample was α͵ǤͻͺͲͻͷΪͲǤͲͲͺͻ͸ ;αͲǤͻͻͻ 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 mg NO 3 --N/L Abs1 - 2 · Abs2 Chapter 2 54 performed with a spectrophotometer (Shimadzu UV-1800) at a wavelength of 640 nm and the quantification was done with help of a calibration curve (Fig. 2.4.). The calibration curve is prepared using a stock solution of 10 mg NH4+L-1, which was diluted in several samples in the range of 0-0.9 mg NH4+-N L-1. Then, the absorbance of each sample was measured in the spectrophotometer and linked to the prepared ammonia concentration. Fig. 2.4. Calibration curve for the determination of the concentration of ammonium. 2.3.3. Solids concentration TSS and VSS were determined according to the methods 2540D and 2540E, respectively, detailed in the Standard Methods (APHA, 2017). For the determination of the TSS, a well-mixed sample of the bioreactors is taken (aiming to yield a residue between 2.5 and 200 mg) and filtered through a weighted glass fiber filter (Whatman, GF/C, 4.7 cm of diameter, 1.2 μm of pore size), which needs to be dried before its use in the oven (around 15 min). The residue retained in the filter is dried in the oven at 105ºC until a constant weight is reached (approximately 2-3 h). The weight increase in the filter represents the TSS content in the sample, as shown in the following equation: ሺȀሻൌ୅ି୆ ୚(5) αͲǤ͹ͶͶͻʹǦ ͲǤͲ͸ͺͳͺ ;αͲǤͻͻͻͺͺ 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 mg NH4+-N/L Absorbance Materials, methods and equipment 55 Where A represents the weight of the oven-dried filter, B the weight of its tare and V the volume of the liquid sample taken from the reactor. The VSS is determined by transferring the dried filters, which allowed to calculate the TSS, to a muffle furnace for calcination at 550ºC for 20-30 min, until a constant weight is reached. The weight loss during ignition corresponds to the volatile fraction (Eq. 6), providing an estimate of the organic matter concentration present in the solid fraction of the bioreactors. ሺȀሻൌ୅ିେ ୚(6) Where A stands for the weight of the oven-dried filter, C for the weight of the filter after calcination at 550ºC and V for the volume of the liquid sample taken from the reactor. 2.3.4. Dissolved oxygen, temperature and pH The pH was measured in the reactors using an electrode (Crison Instruments GLP22, USA) equipped with an automatic compensatory temperature device connected to a pH- meter (GLP-22, Crison Instruments). The electrode was calibrated at room temperature with two standard buffer solutions of pH 7.00 and 4.00. The dissolved oxygen was measured using a probe with a membrane-covered galvanic dissolved oxygen sensor (LDO101) connected to a digital multimeter device (HQ40D, HACH Lange, USA). This same probe was equipped with a thermopar that was used to take temperature measurements. 2.3.5. Organic Micropollutants The determination of OMPs concentration in the liquid phase of the bioreactors begins with a centrifugation step at 3500 rpm for 10 min, while the feeding samples, which lack a solid fraction, do not need this step. Then, the supernatants and the feeding samples are prefiltered (AP4004705, Millipore) and filtered again at 0.45 μm (HAWP04700, Millipore). Lastly, solid-phase extraction (SPE) is performed with 200 mL samples and 60 mg Oasis HLB cartridges (Waters, Milford, MA, USA), as described by Fernandez-Fontaina et al. (2013). The quantification of antibiotics (ERY, ROX, SMX, TMP), neurodrugs (FLX, CBZ, DZP) and hormones (E1, E2, EE2) was performed using an Agilent G1312A liquid chromatography instrument with a binary pump and automatic Chapter 2 62 ୢେ౓ ୢ୲ ൌ୊౟౤ሺ୲ሻ ୚െ୊౥౫౪ሺ୲ሻ ୚െቀୢେ౓ ୢ୲ ቁ୴୭୪ୟ୲୧୪୧୸ୟ୲୧୭୬ െቀୢେ౓ ୢ୲ ቁୱ୭୰୮୲୧୭୬ ൅ቀୢେ౓ ୢ୲ ቁୢୣୱ୭୰୮୲୧୭୬ െ ቀୢେ౓ ୢ୲ ቁୠ୧୭୲୰ୟ୬ୱ୤୭୰୫ୟ୲୧୭୬ (12) ୢେ౏ ୢ୲ ൌ୊౟౤ሺ୲ሻ ୚െ୊౥౫౪ሺ୲ሻ ୚൅ቀୢେ౓ ୢ୲ ቁୱ୭୰୮୲୧୭୬ െቀୢେ౓ ୢ୲ ቁୢୣୱ୭୰୮୲୧୭୬ (13) Where CW(μg L-1) is the dissolved OMPs concentration, Cs(μg L-1) is the OMPs concentration in the solid phase, Fin (μg d-1) is the inlet OMPs flow rate, Fout (μg d-1) is the outlet OMPs flow rate and V (L) is the bioreactor volume. Volatilization is based on the transfer due to an equilibrium process between the liquid and gas phase and depends mostly on the physicochemical properties of the OMPs and the operating conditions of the process. In an activated sludge process, volatilization can occur due to stripping, which is controlled by the air flow, the OMPs concentration and the Henry’s constant, and due to surface volatilization, which usually is not taken into account as a result of its lower relevance (Ltd I C Consultants, 2001). Volatilization can be neglected when a compound has a Henry constant below 10 Pa m3mol-1 and its Hc/Kow (Henry’s constant divided by the octanol-water partition coefficient) is lower than 10-4 (Ltd I C Consultants, 2001). This is the case of the OMPs studied in this thesis (Table 2.1.). Sorption is based on the transfer due to an equilibrium process between the liquid and solid phase, including dissolved and colloidal organic matter, where OMPs simultaneously undergo sorption and desorption phenomena (Barret et al., 2010). Equilibrium is assumed to be reached generally in a matter of hours and depends on the OMPs physicochemical properties (including hydrophobicity, cation exchange, cation bridging, surface complexation and hydrogen-bonding (Plósz et al., 2010)), sludge quality and operating conditions (Pomiès et al., 2013). Two mechanisms are involved in the sorption process: (i) absorption, occurring due to hydrophobic interactions between the aliphatic and aromatic groups of the OMPs and the lipophilic cell membrane and fat fractions of the microorganisms and (ii) adsorption, happening due to electrostatic interactions of the positively charged groups of the OMPs and the negatively charged surfaces of the microbial population (Sipma et al., 2010). Since OMPs are present at low Materials, methods and equipment 63 concentrations in WWTPs, a linear model is normally assumed to describe this phenomenon (Pomiès et al., 2013), as shown below (Eq. 14 and 15): ቀୢେ౓ ୢ୲ ቁୱ୭୰୮୲୧୭୬ൌୱ୭୰൉୛൉୘ୗୗ (14) ቀୢେ౓ ୢ୲ ቁୢୣୱ୭୰୮୲୧୭୬ൌୢୣୱ୭୰൉ୗ(15) Where ksor is the sorption kinetic constant (L g-1 d-1), kdesor is the desorption kinetic constant (d-1) and XTSS is the total suspended solids concentration (g L-1). The partition coefficient (Kd), shown in Eq. 16, is the ratio between the kinetic constants at equilibrium conditions: ୢൌ୩౩౥౨ ୩ౚ౛౩౥౨ ൌେ౏ େ౓൉ଡ଼౐౏౏ (16) Biotransformation consists of the biological removal of the compounds, which is typically calculated for the dissolved OMPs (Pomiès et al., 2013). Pseudo-first order kinetics are typically used to model the OMPs biotransformation rate (rbiol), represented in Eq. 17 (Schwarzenbach et al., 2003): ቀୢେ౓ ୢ୲ ቁୠ୧୭୲୰ୟ୬ୱ୤୭୰୫ୟ୲୧୭୬ ൌୠ୧୭୪ൌୠ୧୭୪൉୛൉୚ୗୗ (17) The biodegradation kinetic constants were calculated performing mass balances to the batch experiments (Chapter 3), as shown in Eq. 18: ୠ୧୭୪ሺିଵିଵሻൌେ౓బିେ౓౪൉ሺଵା୩ౚ൉ଡ଼౐౏౏ሻ େ౓౪൉ଡ଼౒౏౏൉୲ (18) Where Cw0 (μg L-1) is the dissolved OMPs concentration at time 0 and Cwt (μg L-1) is the dissolved OMPs concentration at time t, where the slope is maximum. The biotransformation yield (Eq. 19) and the specific biotransformation rate (Eq. 20) were calculated to describe the behavior of the OMPs. The yield determines the biotransformation efficiency of the compounds and the biotransformation rate provides information regarding the cometabolic behavior of the OMPs and the performance of the reactor. Chapter 2 64 ሺΨሻൌ୊౟౤ି୊౭౥౫౪ି୊౩౥౫౪ ୊౟౤ ൉ͳͲͲ (19) ሺρ୚ୗୗିଵିଵሻൌ୊౟౤ି୊౭౥౫౪ି୊౩౥౫౪ ଡ଼౒౏౏൉୚ ൉ͳͲͲ (20) Where Fwout (μg d-1) is the outlet dissolved OMPs flow rate and Fsout (μg d-1)is the outlet OMPs flow rate in the solid phase. The outlet OMPs flow rate in the solid phase (Fsout) is calculated as shown in Eq. 21: ୱ୭୳୲ ൌୗԢ൉୘ୗୗିୣ୤୤୪୳ୣ୬୲ ൉ୣ୤୤୪୳ୣ୬୲൅ୗԢ൉୘ୗୗି୮୳୰୥ୣ൉୮୳୰୥ୣ (21) Where CS’ (μg g-1) is the OMPs concentration in the solid phase, Feffluent (L d-1) is the flow of the effluent and Fpurge (L d-1) is the flow of the purge. In Chapter 3, to determine if the yields and specific biotransformation rates of the OMPs were statistically different at the tested primary substrate specific biodegradation rates, R software 3.6.2. was used. The statistical tests were performed at a 5% significance level (p < 0.05). CHAPTER 3 CHAPTER 3. The organic loading rate affects organic micropollutants’ cometabolic biotransformation kinetics under heterotrophic conditions in activated sludge SUMMARY Organic micropollutants (OMPs) are biotransformed cometabolically in activated sludge systems. However, the individual role that heterotrophs play is still not clear and there is still a gap regarding the influence of the heterotrophic activity on the cometabolic biotransformation kinetics and yield of the OMPs. To answer these questions, experiments with increasing primary substrate concentrations were performed under aerobic heterotrophic conditions in a continuous stirred tank reactor operated at several organic loading rates (OLR) with fixed hydraulic retention time. Moreover, the individual kinetic parameters were determined in batch assays with different initial substrate concentrations using the sludges from the continuous reactor. A set of 15 OMPs displaying a variety of physicochemical properties were spiked to the feeding in the ng L-1 -μg L-1 range. Results reveal that the biodegradation of the primary carbon source and the biotransformation of the OMPs occur simultaneously, in evidence of cometabolic behaviour. Moreover, the OMPs biotransformation kinetic constant (kbiol) shows a linear dependence with the OLR of the primary substrate for most compounds, suggesting that the heterotrophic activity affects the OMPs biotransformation kinetics. However, under typical activated sludge systems operating conditions (hydraulic retention times above 8 h), the biotransformation yield would not be significantly affected. This chapter, under the consent of the publisher (Elsevier), was redrafted after: Kennes-Veiga, Gonzalez-Gil, L., Carballa, M., Lema, J. M. (2021). The organic loading rate affects organic micropollutants’ cometabolic biotransformation kinetics under heterotrophic conditions in activated sludge. Water Research, 189, 116587. ISSN: 0043- 1354. Further information can be found in the “List of publications” (page 145). Chapter 3 66 3.1. INTRODUCTION To characterize OMPs biotransformation, a pseudo-first order kinetic model is commonly assumed and the resulting biotransformation kinetic constant (kbiol) largely depends on the physicochemical properties of each particular pollutant (Lema and Suarez, 2017). In fact, Falås et al. (2016) concluded that kbiol varies more depending on the nature of the OMPs than on the experimental conditions. However, kbiol, as a kinetic constant, is also affected by the reactor operating conditions (such as temperature, pH and oxidation reduction potential), the presence and availability of co-substrates and the biochemical versatility of the sludge, as several studies demonstrate (Barceló, 2012; Petrovic et al., 2013). For instance, Alvarino et al. (2016) reported kbiol values of 0.09 and 0.05 L g VSS- 1d-1 for SMX under aerobic heterotrophic and autotrophic denitrifying conditions, respectively; Gulde et al. (2014) reported kbiol increments of one order of magnitude for atenolol at pH 8 compared to pH 6 under activated sludge conditions and Li et al. (2005) reported that E2 increased its kbiol value from 1.8 to 3.3 L g VSS-1 h-1 when increasing the temperature from 20 to 35ºC in batch experiments with activated sludge. Nonetheless, conversely, there are so far no conclusive studies regarding the influence of the OLR and the heterotrophic activity on kbiol. The role of the OLR in the removal of the OMPs is, in fact, a topic of current and general interest in biological systems, but results still do not show a clear trend. For instance, Kora et al. (2020) showed that increasing the OLR improved the removal of 3 out of 5 OMPs spiked in a methanogenic-aerobic moving bed biofilm reactor and Moya- Llamas et al. (2018) found better biodegradations for 6 OMPs at higher OLRs in a UASB reactor coupled to a MBR. On the contrary, in a moving biofilm bed reactor, Abtahi et al. (2018) reported for some OMPs a maximum removal at the highest OLR tested and for others at the lowest OLR and Carneiro et al. (2020) reported a negative impact of increasing OLRs in the biodegradation of citalopram and SMX in anaerobic fix bed biofilm reactors. Finally, differently from these studies, Gonzalez-Gil et al. (2018) found no correlation between variations in the OLR and the biotransformation of most OMPs in methanogenic digesters. The objective of the present chapter is to extend the knowledge behind OMPs biotransformation processes under exclusively aerobic heterotrophic conditions. More specifically, research is focussed on assessing the OMPs cometabolic biotransformation, The OLR affects OMPs cometabolic biotransformation kinetics 67 aiming to determine the relationship between the intensity of the heterotrophic activity and the OMPs biotransformation rate, with a particular focus on evaluating the effect on kbiol. The implications of understanding such linkage could be of great importance to design biological systems able to maximize OMPs removal. To reach these goals, a series of experiments in a continuous stirred tank reactor operated at different OLRs were performed, as well as batch experiments set with different initial COD concentrations. 3.2. MATERIALS AND METHODS An aerated 5 L continuous stirred tank reactor connected to a 2 L settler (section 2.2.1.) set at an HRT of 1 d and a SRT around 5 d was inoculated with activated sludge at a concentration of 1 g L-1. It was operated at four different OLRs (0.2, 0.4, 0.6 and 0.8 g COD L-1 d-1, that resulted in 0.4, 0.6, 1.1 and 2.7 g COD g VSS-1 d-1), selected trying to cover typical conditions of activated sludge WWTPs (Lema and Suarez, 2017; Metcalf & Eddy, 2014). Each experimental stage was maintained for one month and the OLR was changed by varying the organic matter concentration in the feeding while keeping constant the HRT. The feeding was based on acetate as organic carbon source (section 2.1.2.) and contained ERY, ROX, TMP, SMX, IBP, NPX, CBZ, DZP and TCS at concentrations of 10 μg L-1; E1, E2 and EE2 at 1 μg L-1 and ADBI, HHCB and AHTN at 40 μg L-1 (section 2.1.1). Once steady state was reached at each OLR, inlet and outlet OMPs concentrations were measured (solid and liquid phase) by taking three samples from the feeding and the reactor vessel in three consecutive days. Besides, analysis of the influent and effluent of the reactor was performed to determine the typical operational parameters according to the standard analytical methods described in Chapter 2. The soluble COD, ammonium, nitrate and nitrite concentrations, SSV, pH and temperature were measured two to three times per week. Additionally, 4 batch assays using the same group of OMPs were set to analyze their biotransformation kinetics and kbiol. The experiments were aerated and performed at neutral pH, 25ºC, 150 rpm and COD initial concentrations of 0.2, 0.4, 0.6 and 0.8 g L-1 by varying the organic carbon concentration while maintaining the same micro and macronutrients used in the heterotrophic reactor. The inoculum was taken from the continuous reactors working at the respective OLRs, aiming at an initial concentration of 0.80 g VSS L-1. For each batch assay, 18 flasks were prepared, three for each time point to have triplicates (0, 1, 3, 8, 24 and 48 h). At the mentioned times, 3 flasks were taken Chapter 3 68 to determine the concentration of the OMPs in the solid and liquid phases. The same conventional operational parameters as in the continuous reactor were determined at each time point, except for the solids and nitrogen concentrations which were measured at times 0, 24 and 48 h. 3.3. RESULTS AND DISCUSSION 3.3.1. Biotransformation yield Fig. 3.1 shows the biotransformation yield, i.e. the biotransformation percentage, of the selected OMPs at the 4 OLRs in the aerobic heterotrophic reactor. The biotransformation extent varied considerably among OMPs, being their behaviour characterized as (i) low biotransformation (below 20%) for CBZ and DZP, (ii) mediumlow biotransformation (20-50%) for TMP, (iii) medium-high biotransformation (50-80%) for ERY and (iv) high biotransformation (over 80%) for ROX, SMX, IBP, NPX, TCS, ADBI, HHCB, AHTN, E1, E2 and EE2. These results are consistent with literature information (Alvarino et al., 2018b; Fernandez-Fontaina et al., 2016). Fig. 3.1. OMPs biotransformation yield achieved in the aerobic heterotrophic reactor operated with fixed HRT (1 day) at 4 different specific OLRs of the primary substrate. 0 10 20 30 40 50 60 70 80 90 100 ERY ROX SMX TMP IBP NPX TCS ADBI HHCB AHTN E1 E2 EE2 CBZ DZP Biotransformation (%) 0.4 g COD/g VSS·d 0.6 g COD/g VSS·d 1.1 g COD/g VSS·d 2.7 g COD/g VSS·d The OLR affects OMPs cometabolic biotransformation kinetics 69 It can be observed that the biotransformation yield of most OMPs remained constant throughout the different OLRs (biotransformation differences below 10 percentage points and not statistically significant; p > 0.05 ). Only one value for ERY (experiment with 1.1 g COD g VSS-1 d-1) and another for IBP (experiment with 2.7 g COD g VSS-1 d-1), significantly (p < 0.05) deviated (between 15-20 percentage points) from the values of the other three conditions. These events for ERY and IBP cannot be attributed to a specific trend or behaviour. Therefore, overall, the results of Fig. 3.1 indicate that increasing the aerobic heterotrophic cometabolism did not affect the OMPs biotransformation yield under the conditions tested. Several hypotheses could explain this behaviour. The first one is that the enzymes involved in the OMPs biotransformation might not be increasingly stimulated at higher OLRs. According to Stadman (1970), in catabolic pathways, enzymes can be classified as “constitutive” enzymes, whose concentration is independent of the presence of their substrates, and “inducible” enzymes, that are produced when their immediate substrates or suitable derivatives are present. If the enzymes involved in the biotransformation of the OMPs belong to the first class and were not increasingly stimulated, the yield could have remained stable even at higher OLRs. A second hypothesis is that the maximum cometabolic rate towards OMPs is already reached at the lowest OLR and, therefore, increasing the degradation rate of the primary substrate (acetate) does not have an effect on the OMPs. In this regard, it is commonly assumed that the oxidation rate of a non-growth substrate (OMPs) should always be linked to the oxidation rate of a growth substrate (acetate) proportionally; however, some studies have shown that this is not necessarily always the case (Criddle, 1993; M. H. Kim et al., 2020). Thirdly, the OMPs may have already achieved their biotransformation limits due to thermodynamic constraints. This event could be caused by reversibility of the biological reactions, leading to a chemical equilibrium between the parent compound and the TPs, as previously suggested (Gonzalez-Gil et al., 2019a). Lastly, the primary metabolism may increase the specific biotransformation rate, but the HRT may have been high enough to hide this effect, showing the same biotransformation yield in all cases. In common biological treatments, as in activated sludge, the optimal HRT for OMPs removal is 24 h or longer (Boonnorat et al., 2019). Nevertheless, it has been proven that lower times may be sufficient to achieve the maximum biotransformation extent for several compounds (Boonnorat et al., 2019; Ejhed et al., 2018). Overall, the results shown in Fig. 3.1 are insufficient to determine which of these four hypotheses is more likely, being necessary to evaluate the behavior of the biotransformation rate. Chapter 3 70 3.3.2. Specific biotransformation rate In Fig. 3.2, it can be observed the direct relationship between the OMPs biotransformation rate and the specific sludge activity: the higher the specific biodegradation rate of the growth substrate, the higher the OMPs specific biotransformation rate. The dependence of the biotransformation rate of the secondary substrates (OMPs) with the biodegradation of the primary substrate (acetate) is clear proof of cometabolism (Lema and Suarez, 2017). When the concentration of the primary substrate is increased (i.e., a higher OLR is applied), a higher microbial activity of the biomass is achieved. Hence, if an increment of the OMPs biotransformation rate also occurs, it indicates that the OMPs are been cometabolized by the same enzymes involved in the metabolism of the growth substrate. In Fig. 3.2, the differences in the specific biotransformation rates were statistically significant (p < 0.05) for all experimental conditions and all OMPs tested. A) 0.0 0.5 1.0 1.5 2.0 E1 E2 EE2 Biotransformation rate ( P g OMP biotransformed /g VSS·d) 0.4 g COD/g VSS·d 0.6 g COD/g VSS·d 1.1 g COD/g VSS·d 2.7 g COD/g VSS·d The OLR affects OMPs cometabolic biotransformation kinetics 71 B) C) Fig. 3.2. Specific OMPs biotransformation rate (Pg OMP g VSS-1 d-1) at 4 specific biodegradation rates of the primary substrate (g COD g VSS-1 d-1). A) OMPs fed at 1 Pg L-1 d-1. B) OMPs fed at 40 Pg L-1 d-1. C) OMPs fed at 10 Pg L-1 d-1. 0 10 20 30 40 50 60 ADBI HHCB AHTN Biotransformation rate ( P g OMP biotransformed /g VSS·d) 0.4 g COD/g VSS·d 0.6 g COD/g VSS·d 1.1 g COD/g VSS·d 2.7 g COD/g VSS·d 0 5 10 15 20 25 30 ERY ROX SMX TMP IBP NPX TCS Biotransformation rate ( P g OMP biotransformed /g VSS·d) 0.4 g COD/g VSS·d 0.6 g COD/g VSS·d 1.1 g COD/g VSS·d 2.7 g COD/g VSS·d Chapter 3 linearly. The exceptions of HHCB and ADBI could indicate that they would have required lower or higher microbial activities, proving that kbiol is influenced both by the microbial activity and the nature of the compound. To better understand the influence of the microbial activity on the biotransformation of the OMPs, Table 3.1 shows an estimation of the biotransformation rates that could be achieved for ROX, NPX, SMX and HHCB in a real activated sludge system designed to only remove organic matter, as well as the HRT that these OMPs would require for complete biotransformation according to the experimental kbiol values (Fig. 3.4). For instance, in the case of NPX, assuming an inlet concentration of 10 Pg L-1, a kbiol equivalent to 0.7 L g VSS-1 d-1 would provide a biotransformation rate of 14 Pg NPX L-1 d-1 and an HRT of 17 h would be required. On the other hand, a kbiol of 3.8 L g VSS-1 d- 1would lead to a biotransformation rate of 76 Pg NPX L-1 d-1 and an HRT of 3 h could be sufficient for total NPX removal. The results indicate that higher heterotrophic activities could considerably reduce HRT requirements thanks to improved OMPs specific biotransformation rates. Moreover, except for DZP and CBZ, which are recalcitrant under heterotrophic conditions, all the OMPs would be almost completely removed after 8 h (Table 3.1). Accordingly, promoting the heterotrophic activity with higher OLRs would not have significant effects on the biotransformation yield at typical HRTs of activated sludge systems, supporting the findings observed in Fig. 3.1. 78 The OLR affects OMPs cometabolic biotransformation kinetics 79 Table 3.1. Biotransformation rate and HRT required for full biotransformation of ROX, NPX, SMX and HHCB in a real WWTP based on the kbiol values obtained in the batch assays. The XVSS value used is a typical solids concentration in WWTPs (Metcalf & Eddy, 2014), and the C w value is in the range of typical influent OMPs concentrations in WWTPs (Luo et al., 2014; Petrie et al., 2014). OMPs Specific biotr. rate (g COD g VSS -1 d-1) kbiol (L g VSS-1 d-1) X VSS (g L - 1) CW (ߤ ߤ OMP L - 1) Biotransformation rate ( ߤ g OMP L-1 d-1) HRT required for 100% biotransformation (h) ROX 0.4 1.6 210 32 7.6 0.6 2.4 49 4.9 1.1 3.4 68 3.5 2.7 4.0 81 3.0 NPX 0.4 0.7 210 14 17.1 0.6 2.8 56 4.3 1.1 3.4 68 3.5 2.7 3.8 76 3.2 SMX 0.4 1.0 210 20 12.0 0.6 2.0 40 6.0 1.1 2.6 52 4.6 2.7 2.9 58 4.1 HHCB 0.4 5.0 210 100 2.4 0.6 3.9 78 3.1 1.1 6.7 134 1.8 2.7 5.2 104 2.3 Chapter 3 3.4. CONCLUSION In this chapter, experimental data evidenced that higher OLRs lead to higher OMPs biotransformation rates and that the removal of the organic matter and the OMPs occurs simultaneously, proving cometabolism as the main mechanism behind the biotransformation of OMPs under aerobic heterotrophic conditions. The enhancement of the OMPs biotransformation rate occurred due to increases in their respective kbiol values, showing that the biotransformation kinetic constant is not only dependent on the compound and the environmental conditions, but also on the intensity of the metabolic activity. Besides, the influence of the heterotrophs in kbiol is proved to be compound dependent. On the other hand, the cometabolic trend does not necessarily involve an improvement in the biotransformation yield of the compounds providing that a sufficient HRT is applied. In fact, in this work, it is shown that the effectiveness did not improve at higher specific degradation rates of the primary substrate. Hence, the variation of the WWTP organic load, the management of the organic substrates and the control of the microbial activity appear as key parameters governing OMPs biotransformation. 80 CHAPTER 4 CHAPTER 4. Heterotrophic enzymatic biotransformations of organic micropollutants in activated sludge SUMMARY While heterotrophic microorganisms constitute the major fraction of activated sludge biomass, the role of heterotrophs in the biotransformation of organic micropollutants (OMPs) has not been fully elucidated. Yet, such knowledge is essential, particularly when conceiving novel wastewater treatment plants based on a two-stage process including an A-stage under heterotrophic conditions and a B-stage based on anammox activity. Biotransformation of OMPs in activated sludge is thought to mostly occur cometabolically thanks to the action of low specificity enzymes involved in the metabolism of the primary substrates. For a better understanding of the process, it is important to determine such enzymatic activities and the underlying mechanisms involved in OMPs biotransformation. This task has proven to be difficult due to the lack of information about the enzymatic processes and the complexity of the biological systems present in activated sludge. In this chapter, a continuous aerobic heterotrophic reactor spiked with 20 OMPs at environmental concentrations was operated to (i) assess the potential of heterotrophs during the cometabolic biotransformation of OMPs, (ii) identify biotransformation reactions catalysed by aerobic heterotrophs and (iii) predict possible heterotrophic enzymatic activities responsible for such biotransformations. Contradicting previous reports on the dominant role of nitrifiers in OMPs removal during activated sludge treatment, the heterotrophic population proved its capacity to biotransform the OMPs to extents equivalent to reported values in nitrifying activated sludge plants. Besides, 12 transformation products potentially formed through the activity of several enzymes present in heterotrophs, including monooxygenases, dioxygenases, hydrolases and transferases, were identified. This chapter, under the consent of the publisher (Elsevier), was redrafted after: Kennes-Veiga, D. M., Vogler, B., Fenner, K., Carballa, M., Lema, J. M. (2021). Heterotrophic enzymatic biotransformation of organic micropollutants in activated sludge. Science of the Total Environment, 780, 146564. ISSN: 0048-9697. Further information can be found in the “List of publications” (page 145). Chapter 4 82 4.1. INTRODUCTION Since OMPs are frequently not fully mineralized during cometabolic biotransformation, the assessment of the TPs is also important because they pose the potential to be as or even more toxic than their parent compounds to the ecosystem (Berkner and Thierbach, 2014; Celiz et al., 2009; Gulde et al., 2016). Besides, TPs identification can be a very helpful tool to determine the reactions occurring in a specific environment and provide hints about the enzymatic activities catalysing OMPs biotransformation. For instance, following this approach, Gulde et al. (2016) demonstrated that N-acyltransferases could be involved in the biotransformation of several amine-containing OMPs in activated sludge. In this chapter, we aim to address the contribution of heterotrophs to the biotransformation of OMPs in activated sludge processes and, particularly, their capacity to biotransform a range of OMPs. Besides, we attempt to detect several TPs as a tool to identify key OMP biotransformation reactions catalysed by aerobic heterotrophs and decipher possible enzymatic activities carrying out such biotransformations. To this end, we evaluated the biotransformation of a set of OMPs with different physicochemical properties in an aerobic heterotrophic reactor and used liquid chromatography coupled to high-resolution mass spectrometry to identify TPs. 4.2. MATERIALS AND METHODS A 5 L continuously stirred lab-scale reactor connected to a 2 L settler (section 2.2.1.) was operated at 25ºC with an OLR of 0.6 g COD L-1 d-1, HRT of 1 d and SRT of approximately 5 d to determine the key enzymatic activities involved in the biotransformation of several OMPs. The reactor was inoculated with activated sludge at a concentration of 1 g L-1 and the experimental stage was maintained for one month. The feeding included ATU to avoid the development of nitrifying activity and was based on acetate as organic carbon source (section 2.1.2.). Besides, 20 OMPs (section 2.1.1.) were spiked at concentrations of 10 μg L-1 for the pharmaceuticals, 1 μg L-1 for the hormones and 40 μg L-1 for the fragrances. Once steady state was reached, inlet and outlet OMPs and TPs concentrations were measured (solid and liquid phase) by taking triplicate samples from the feeding and the reactor vessel in three consecutive days. Besides, analysis of the influent and effluent of Heterotrophic enzymatic biotransformations of OMPs 83 the reactor was performed to determine the typical operational parameters according to the standard analytical methods described in Chapter 2. The soluble COD, ammonium, nitrate and nitrite concentrations, TSS, SSV, pH and temperature were measured two to three times per week. 4.3. RESULTS AND DISCUSSION 4.3.1. Reactor performance The reactor was operated for 28 days and steady-state conditions were achieved after one week. During experimentation, neutral pH was maintained to maximize the activity of the microorganisms as well as to avoid any effects of varying pH levels on OMPs removal. Although frequently receiving little attention, pH is a key parameter affecting OMPs biotransformation. Its increase by just one pH unit can considerably promote the removal of OMPs with cationic-neutral specification and hinder that of compounds with neutral-anionic speciation, as shown by Gulde et al. (2014). Biomass concentration remained quite stable at 1.4 g L-1 by controlling the purge (Fig. 4.1), performed almost daily from the settler, and the control of the recirculation, set at a ratio of 100% of the influent flow (Eq. 4). The source of organic matter was acetate, which was fully consumed (95-100%), leading to effluent COD concentrations consistently below 0.03 g L-1 (Fig. 4.1). Besides, operation occurred in the absence of nitrification (Fig. 4.1), indicating that all the nitrogen consumed was used only for microbial growth. Among the non-consumed nitrogen, some soluble non-degradable organic nitrogen from endogenous respiration may have been present. The net waste activated sludge produced was estimated at 0.15 g VSS L-1, including the active heterotrophic biomass formed daily and the cell debris, which accounts for 10- 20% of the decayed biomass and cannot be degraded due to their extremely low hydrolysis kinetics (Liu and Wang, 2015). Moreover, there was also an endogenous decay rate of biomass in the reactor, proportional to the active biomass concentration (XVSS), that accounted for 0.14 g VSS L-1 d-1. The endogenous decay represents the cell biomass loss and includes an internal decay (cell level), involving the oxidation of stored substrates to produce energy for cell maintenance, and an external decay (community level), such as cell death or predation by higher organisms (Liu and Wang, 2015). Our reactor presented mostly aerobic heterotrophic bacterial strains, as well as higher life- Chapter 4 forms, such as rotifers and protozoans, based on microscopic observation. The high life forms have also been shown to participate in the removal of OMPs, as reported by Gulde et al. (2018) for protozoa, which seem to be involved in the ion trapping of aminecontaining compounds and the hydrolysis of select esters and phenylurea compounds. 84 Heterotrophic enzymatic biotransformations of OMPs 85 A) B) 0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 0 5 10 15 20 25 30 Purge (L) Volatile suspended solids (g/L) Time (d) SSV (g/L) Purge (L) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 20 40 60 80 100 0 5 10 15 20 25 30 COD concentration (g/L) COD removal (%) Time (d) COD removal (%) Influent COD (g/L) Effluent COD (g/L) Chapter 4 86 C) Fig. 4.1. Operational parameters obtained in the continuous aerobic heterotrophic reactor. A) VSS concentration and purge performed. B) COD removal along with COD influent and effluent concentrations. C) Influent N-NH4+and effluent N-NH4+, N-NO3-and N- NO2-concentrations. 4.3.2. Heterotrophic OMPs biotransformation in activated sludge systems Fig. 4.2 shows the fate of the selected OMPs in the continuous heterotrophic reactor. Sorption appears to be minimal for most compounds (<10%), except for FLX, for which it accounts for 15-20% of the total mass balance. Results show that 17 out of the 20 OMPs were highly removed (above 80%) due to heterotrophic activity and only 3 compounds were slightly (TMP) or not removed at all (CBZ, DZP). These results agree with the biotransformation observed in activated sludge units (Alvarino et al., 2014; Luo et al., 2014; Petrie et al., 2014), except for DCF, for which we obtain higher values than under typical activated sludge conditions (Alvarino et al., 2014; Fernandez-Fontaina et al., 2016). Nonetheless, in other biological treatments, high DCF efficiencies have also been achieved, as in a hybrid biofilm-activated sludge process (Jewell et al., 2016b) and in a nitrifying moving bed biofilm reactor (Torresi et al., 2016b). In our experiments, DCF 0 10 20 30 40 0 5 10 15 20 25 30 Concentration (mg/L) Time (d) Influent N-NH4+ (mg/L) Effluent N-NH4+ (mg/L) Effluent N-NO2- (mg/L) Effluent N-NO3- (mg/L) Heterotrophic enzymatic biotransformations of OMPs 87 was biotransformed rather extensively, i.e., at 80% (Fig. 4.2), which could indicate the presence of certain heterotrophic genera capable of biotransforming it more efficiently. In this regard, Nguyen et al. (2019) showed that under exposure to a primary carbon substrate and DCF, certain genera of activated sludge bacteria can significantly increase their abundance, suggesting that they might gain a competitive advantage from its cometabolic removal. Fig. 4.2. Fate of the OMPs in the continuous aerobic heterotrophic reactor operated with fixed HRT (1 d) and OLR (0.6 g COD L-1 d-1). To understand the variability in OMPs biotransformation reported in the literature, Fig. 4.3 compares the biotransformation performance of our aerobic heterotrophic reactor with that reported under nitrifying conditions exclusively (systems not fed with organic carbon or that have inhibited the heterotrophic activity) and in activated sludge systems presenting both heterotrophic and nitrifying activities. Firstly, it can be observed that heterotrophs reach biotransformations similar to the values of the activated sludge plants, indicating that they can greatly contribute to the removal of the OMPs, and that, for certain compounds, there may not be the need to set up WWTPs with nitrifying activities. There is a limited number of studies evaluating the biotransformation of the selected OMPs with exclusively nitrifying populations. In such works, they have also reported 0 10 20 30 40 50 60 70 80 90 100 OMPs fate (%) Biotransformation Sorption Effluent Chapter 4 94 Table 4.1. Summary of the suggested TPs structures found in the aerobic heterotrophic reactor, including the confidence level, atom changes relative to the parent OMPs and biotransformation reactions and enzymes likely involved in the process. The TP structures placed in brackets are only suggestions for possible structures. The functional groups bonded with dotted lines could be placed somewhere else in the molecule. OMPs TP Level Atom chang e Reaction Candidate enzymes ADBI 3 + O Hydroxylation EC 1.14.- EC 1.13.- ADBI 3 + 2O - 2H Hydroxylation Additional oxidation EC 1.14.- EC 1.13.- EC 1.1. - HHCB 4 + 2O Hydroxylation EC 1.14.- EC 1.13.- E1 4 + 2O Hydroxylation EC 1.14.- EC 1.13.- E2 3 + O Hydroxylation EC 1.14.- EC 1.13.- Heterotrophic enzymatic biotransformations of OMPs 95 DCF 3 + O Hydroxylation EC 1.14.- EC 1.13.- NP 3 + 3O - 2H Hydroxylation Additional oxidation EC 1.14.- EC 1.13.- EC 1.1. - TMP 3 + O - N, H Hydrolysis EC 3.5.99.- DZP 3 - C, 2H Demethylation EC 1.14.- FLX 3 + 2O - C, N, 5H Demethylation Deamination Oxidation EC 1.14.- EC 1.4. - EC 1.2.1.- FLX 2 + 4C, 3O, 4H Conjugation EC 2.3. - SMX 3 + 7C, 4N, 2O, 4H Conjugation EC 2.5. - EC 3.5.4.11 Chapter 4 96 4.4. CONCLUSION This chapter highlights the relevant role of aerobic heterotrophs in the cometabolic biotransformation of OMPs, speaking to their capacity to contribute to the overall removal of OMPs in activated sludge plants and questioning the requirement of maintaining a nitrifying activity. In fact, this study shows that heterotrophs can achieve similar extents of biotransformation for the selected OMPs as those reported in several literature studies with activated sludge (with nitrifying, denitrifying and heterotrophic activities) and purely nitrifying systems. Based on the TPs produced, it was possible to identify the main reactions involved in OMPs biotransformation under aerobic heterotrophic conditions, which are oxidation (hydroxylation, dehydrogenation, deamination and demethylation), hydrolysis and conjugation routes. An overall analysis of all results allows selecting mono- and dioxygenases, dehydrogenases, hydrolases and transferases as some of the main enzymatic activities likely responsible for OMPs biotransformation under heterotrophic conditions. Thus, this chapter highlights the relevant contribution of heterotrophs to OMPs removal, which are gaining importance in the conception of new WWTPs, and deepens the knowledge on their biotransformation mechanisms. CHAPTER 5 CHAPTER 5. Sulfamethoxazole triggers specific enzymatic activities under aerobic heterotrophic conditions: A metaproteomic approach SUMMARY Wastewater treatment plants (WWTPs) are a potential hotspot for the development of antibiotic resistance genes, creating the need to study the fate of antibiotics and the evolution of the microbiota when exposed to these compounds. This chapter aims at determining whether the presence of sulfamethoxazole (SMX), even at low concentrations, modifies the microbial structure and enzymatic expression of a heterotrophic activated sludge system. To that end, a combined approach including metaproteomic, genomic and transformation product analyses was followed. Results revealed the formation of the metabolites N4-acetyl-SMX and 2,4(1H,3H)- pteridinedione-SMX, pointing out the presence of several biotransformation pathways. Besides, when exposed to increasing SMX concentrations, five bacterial families belonging to the phylum Proteobacteria followed characteristic trends, the genus Corynebacterium increased its abundance and five enzymes involved in its central metabolism showed a differential expression, suggesting their relevance to mitigate SMX risks. Overall, this chapter confirms the potential of metaproteomic techniques to unravel organic micropollutants’ biotransformation mechanisms in WWTPs and elucidate the bacterial and enzymatic key players. This chapter was redrafted after the following under-review publication: Kennes-Veiga, D. M., Trueba-Santiso, A., Gallardo-Garay, V., Lema, J. M., Carballa, M. Sulfamethoxazole triggers specific enzymatic activities under aerobic heterotrophic conditions: A metaproteomic approach. Environmental Science and Technology. ISSN: 1520-5851. Under review. Further information can be found in the “List of publications” (page 145). Chapter 5 98 5.1. INTRODUCTION The presence of antibiotics in WWTPs poses a serious environmental and health risk due to the development of antibiotic-resistant genes and bacteria (Pazda et al., 2019; Wang et al., 2021; World Health Organization, 2021). Sulfonamides are a group of antibiotics particularly important due to their intensive utilization worldwide and, among them, SMX is the most broadly consumed one (Achermann et al., 2018a; Carvalho and Santos, 2016). In 2020, due to the ongoing growing concern, SMX was included in the “Surface Water Watch List” published by the Water Framework Directive of the European Union to monitor and gather data about its potential risks to the aquatic environment (European Commission, 2020). The knowledge of the enzymes and microorganisms involved in SMX biotransformation in activated sludge systems is still very limited. However, metaproteomics offers a suitable approach to solve these questions since it provides a global view of the proteins expressed by a microbial community at a specific moment (Rodríguez-Valera, 2004), allowing to study biological processes in their native environment while avoiding the time-consuming labor of isolating microorganisms (Lacerda and Reardon, 2009; Yuqiu Wang et al., 2020). The goal of this chapter was to obtain a better insight into the underlying mechanisms involved in the aerobic heterotrophic biotransformation of SMX. Specifically, it is explored how SMX concentrations affect the biotransformation capacity, taxonomic composition, and enzymatic expression of a heterotrophic activated sludge microbial community. For that purpose, a combination of metaproteomic, genomic and TPs analyses was applied. 5.2. MATERIALS AND METHODS A 2 L continuously stirred lab-scale reactor, connected to a 1 L settler (section 2.2.1.), was operated in three different stages. Firstly, the bioreactor was fed without SMX for two months (control stage), then with 50 μg L-1 of SMX for 45 days (stage I) and finally with 1000 μg L-1 of SMX for another 45 days (stage II). The feeding included ATU to avoid the development of nitrifying activity, was based on acetate as organic carbon source (section 2.1.2.) and the reactor was operated with an OLR of 0.5 g COD L-1 d-1, an HRT of 1 d and a SRT of approximately 5 d. Once steady state was reached at each SMX triggers specific enzymatic activities 99 stage, inlet and outlet SMX concentrations (section 2.3.5.) were measured by taking three samples from the feeding and the reactor vessel in three consecutive days. Besides, at the end of each stage, triplicate samples for TPs and 16S rRNA gene sequencing analyses were taken from the reactor (sections 2.3.6. and 2.3.7.). Furthermore, analyses of the influent and effluent of the reactor were performed to determine the typical operational parameters according to the standard analytical methods described in Chapter 2. The soluble COD, ammonium, nitrate and nitrite concentrations, SSV, pH and temperature were measured two to three times per week. Additionally, a total of 18 sequential batch reactors (section 2.2.2.) divided into six groups of triplicates were spiked with SMX to obtain the following concentrations: 0 (control), 50, 250, 500, 1000 and 2000 μg L-1. The reactors were inoculated with activated sludge at a concentration of 1 g L-1, operated at 25ºC and the feeding was also based on acetate as organic carbon source (section 2.1.2.). The experiment lasted 25 d and was performed under sterile conditions. The content of the reactors was centrifuged daily to separate the biomass from the supernatant, which was removed before adding new feed together with the SMX spike. The operation was monitored by measuring the typical parameters according to the standard analytical methods described in Chapter 2. TSS, VSS and ammonium, nitrate, nitrite and oxygen concentrations were determined weekly, while the acetate concentration and pH, daily. Besides, SMX and TPs concentrations were determined on days 2, 17 and 25. Finally, samples for proteomic (days 1 and 25, section 2.3.8.) and 16S rRNA gene sequencing (day 25) analyses were taken to determine the key enzymatic and microbial activities involved in the biotransformation of SMX. 5.3. RESULTS AND DISCUSSION 5.3.1. SMX biotransformation under aerobic heterotrophic conditions In the continuous bioreactor, the presence of SMX did not affect the consumption of the primary carbon source (acetate), possibly due to one of the following reasons: i) the highest SMX concentrations evaluated are still below inhibitory levels and ii) the role of specific bacteria involved in the biotransformation of SMX mitigated the potential negative effects of the antibiotic over other microorganisms. The biotransformation efficiency of SMX was independent of its concentration in the feeding, being just slightly lower in stage I (70 %) than in stage II (80 %). As a consequence, the specific Chapter 5 100 biotransformation rate increased considerably between the two stages, varying from 0.8 to 10.6 μg OMP g-1 VSS h-1. Similarly, in the sequential batch reactors, even the highest SMX concentrations did not affect the biodegradation of acetate, whose removal remained constant at 140 mg L-1 h-1. The average SMX biotransformation ranged between 62 and 78 % depending on the spiked concentration (Table 5.1) and showed clear characteristic trends. Firstly, regardless of the SMX influent concentrations, the biotransformation yield increased from day 2 to day 17 and decreased on day 25 to values even lower than those observed on day 2. Secondly, on days 2 and 25, lower initial SMX concentrations lead to higher biotransformation yields, while this trend was not observed on day 17. The improved biotransformation on day 17 compared to day 2, as well as the reduced biotransformation yield of SMX on day 2 at higher initial concentrations can be attributed to the acclimation phase of the microorganisms capable of biotransforming SMX. Their higher abundance on day 17 might have allowed reaching a constant biotransformation extent in the 80-90 % range in all bioreactors. This agrees with the findings of Li et al. (2016), who after an extended lag phase proportional to the doses amended, observed biotransformation of multiple antibiotics by bacteria from different genera. Nonetheless, the decreased biotransformation on day 25, intensified at higher SMX concentrations, might be associated with the accumulation of TPs that limit the biotransformation of the parent compound due to reversibility events (Gonzalez-Gil et al., 2019a), thermodynamic limitations (Gonzalez-Gil et al., 2018a) or by exerting toxicity over a certain concentration (Cao et al., 2019; Pérez et al., 2005), thus outweighing the increased presence of SMX degraders. These results highlight the capacity of the aerobic heterotrophic population to extensively biotransform SMX, following the results of previous works (Fernandez- Fontaina et al., 2016; Kennes-Veiga et al., 2020; Majewsky et al., 2011). In fact, heterotrophs could be more favorable to biotransform SMX than nitrifying microorganisms (Fernandez-Fontaina et al., 2016; Tran et al., 2013), suggesting that stimulating the heterotrophic activity in activated sludge systems may positively influence its biotransformation. SMX triggers specific enzymatic activities 101 Table 5.1. SMX biotransformation yield in the sequential batch reactors throughout the experiment. Influent SMX concentration (μg/L) Biotransformation (%) Day 2 Day 17 Day 25 Average 50 80 86 68 78 r 8 250 77 87 63 76 r 10 500 72 79 60 70 r 8 1000 70 79 36 62 r 19 2000 63 86 43 64 r 18 5.3.2. TPs identification under aerobic heterotrophic conditions Two SMX TPs were detected in the continuous bioreactor: 2,4(1H,3H)- pteridinedione-SMX (PtO-SMX) and N4-acetyl-SMX. The latter was only observed during stage II, which can be attributed to two reasons: (i) N4-acetyl-SMX may have been also present during stage I, but at a concentration below the limit of identification, or (ii) additional biotransformation routes may have appeared at higher SMX doses due to variations in the concentrations of the metabolites that affected the metabolic flux and pathway selection (Wegner et al., 2015). Actually, OMPs influent concentration is recognized as a key parameter affecting biotransformation yields, rates and pathways (Onesios-Barry et al., 2014; Rios-Miguel et al., 2021; Yuxin Wang et al., 2020). For instance, Jewell et al. (2016) observed variations in the biotransformation pathway and TPs formation of TMP depending on the initial spiked concentration. Differently, in the sequential batch reactors, only PtO-SMX was detected. It was observed in all reactors spiked with SMX at all sampling points, except for the samples taken on day 2 from the reactor fed with 50 μg L-1 of SMX, which is attributed to its expected low concentration in the sample. Multiple SMX TPs formed through a wide range of reactions such as hydroxylation, acetylation, deamination or nitration have been previously reported (Majewsky et al., 2014). PtO-SMX has been detected before in WWTPs effluents and lab experiments using activated sludge (Achermann et al., 2018a). Its formation occurs through the pterinconjugation pathway when sulfonamides interact with dihydropteroate synthetase, hindering folic acid synthesis through competitive inhibition. The pathway begins with SMX conjugation, is followed by oxidation to pterin-SMX and continues with a Chapter 5 102 hydrolysis step to form PtO-SMX thanks to the action of pterin deaminase (EC 3.5.4.11). PtO-SMX can be further broken down through various steps possibly involving oxidation and decarboxylation reactions (Achermann et al., 2018a). Most TPs formed in the sulfonamide pterin-conjugation pathway are modified at the para amino group and still possess the sulfanilamide toxicophore, requiring a detailed risk assessment due to their antibiotic activity, capacity to promote antibiotic resistance and ability to be backtransformed to SMX (Majewsky et al., 2015, 2014; Yun et al., 2012). N4-acetyl-SMX is formed through N-acylation of SMX, likely catalyzed by N- arylamine acetyltransferase (EC 2.3.1.5) (Fischer and Majewsky, 2014; X. Li et al., 2016; Tang et al., 2018). It represents the main SMX human metabolite and has been often detected in WWTPs effluents (Achermann et al., 2018a; Majewsky et al., 2014; L. J. Zhou et al., 2019). Besides, it possesses antibacterial activity and exerts toxicity, although at reduced levels compared to SMX (Majewsky et al., 2014), and it can be rapidly backtransformed to the parent compound (Achermann et al., 2018a; Radke et al., 2009), hindering the accurate assessment of its environmental relevance. 5.3.3. Impact of SMX on microbial community structure The results obtained in the continuous bioreactor by DNA metabarcoding at the family level show a clear shift in the bacterial community related to the presence of SMX and its initial concentration in the feeding (Fig. 5.1.). This finding agrees with previous studies pointing out that the microbial structure can change under the selection pressure of many OMPs (Aguilar-Romero et al., 2020; Vasiliadou et al., 2018; G. Zhou et al., 2019). However, as previously mentioned and different from other studies (Jiang et al., 2017), the consumption of the main carbon source was not affected by these changes. Additionally, the variations in the microbial community between stages did not translate into changes in the SMX removal efficiency, which was maintained in the 70-80 % range. Overall, the highlight is the characteristic trends followed by five bacterial families (Rhodobacteraceae,Comamonadaceae,Xanthobacteraceae,Devosiaceae and Hyphomicrobiaceae) belonging to the phylum Proteobacteria. Such phylum has been described before as one of the most dominant and prevalent in aerobic activated sludges and it has been linked to OMPs removal (Bains et al., 2019; Balcom et al., 2016; Gallardo- Altamirano et al., 2019; Vasiliadou et al., 2018). Commamodaceae was the only bacterial family negatively affected by the presence of SMX at all stages, varying its relative abundance from 17.5 % in the control stage to SMX triggers specific enzymatic activities 103 3.9 and 1.1 % during stages I and II, respectively. Interestingly, Commamodaceae is the only Betaproteobacteria of the 5 families, since the others belong to the Alphaproteobacteria class, which has been shown to thrive in activated sludge systems in presence of several OMPs (Davids et al., 2017). Hyphomicrobiaceae was positively affected when SMX was added to the reactor. It was not detected in the control stage but then its abundance increased to 0.6 and 4.0 % with SMX during stages I and II, respectively. In previous studies, its abundance was related to improved OMPs removal (Coll et al., 2020; Zhang et al., 2020), although in this work it was not observed a variation in the removal efficiency of SMX between stages. Devosiaceae showed a fast and stable adaptation to SMX. It was positively affected by the presence of SMX at 50 μg L-1, increasing its abundance from 0.6 % in the control stage to 10.6 %, which then slightly dropped when SMX was spiked at 1000 μg L-1. The sum of the relative abundances of Rhodobacteraceae and Xanthobacteraceae remained approximately constant throughout the three experimental stages, suggesting a balance between them whereby when one increased, the other decreased, and vice versa. Rhodobacteraceae, which was initially highly abundant (70 %) and has previously been linked to improved removal efficiencies of several OMPs (Gallardo-Altamirano et al., 2019), showed a very particular trend. When SMX was spiked at 50 μg L-1, its abundance decreased approximately by half, but then, when the concentration was changed to 1000 μg L-1, the relative abundance reached values close to those of the control stage. It is hypothesized that firstly it was outcompeted by other microorganisms that had a faster adaptation to the introduction of SMX. Nonetheless, the capacity of Rhodobacteraceae to withstand the highest concentrations of SMX may have allowed them to increase their abundance again. Conversely, Xanthobacteraceae was positively affected by the presence of SMX at 50 μg L-1 but its relative abundance dropped when SMX was spiked at 1000 μg L-1. This suggests that it thrived when the SMX concentration was low but that it could have been outcompeted when SMX was added at much higher concentrations. In this sense, it has been reported before that lower OMPs concentrations may improve microbial richness and diversity (Jiang et al., 2017; Vasiliadou et al., 2018; Zhang et al., 2020; G. Zhou et al., 2019). However, higher concentrations may cause the opposite, as observed in hospital WWTPs, which typically have high pharmaceutical concentrations and reduced bacterial diversity in comparison to urban wastewaters (Vasiliadou et al., 2018). Chapter 5 110 5.4. CONCLUSION This chapter highlights the capacity of the heterotrophic sludge to biotransform SMX while still maintaining the high biodegradation levels of the primary substrate. However, under exposure to increasing SMX concentrations, the composition of the heterotrophic microbiota changed. Firstly, Rhodobacteraceae,Comamonadaceae,Xanthobacteraceae, Devosiaceae and Hyphomicrobiaceae showed characteristic trends in the continuous bioreactors. Secondly, the abundance of Corynebacterium varied in the sequential batch reactors, pointing towards its relevant role to maintain the fitness of the microbial community thanks to the action of some enzymes of its central metabolism. Thus, new insights to understand the SMX biotransformation mechanisms and decipher the key microorganisms and enzymes under aerobic heterotrophic conditions are provided. Furthermore, the advantages and usefulness of metaproteomics to study OMPs biotransformation mechanisms are proved, providing arguments to the extensive debate raised in the last few years about the benefits and limitations of omic techniques in environmental engineering applications (Achermann et al., 2020; Fenner et al., 2021; David R. Johnson et al., 2015b; Kennes-Veiga et al., 2021a). However, some enzymes that should have been present in the sludge could not be detected, suggesting that further technical developments to lower the identification and quantification thresholds are still required. CHAPTER 6 CHAPTER 6. GENERAL DISCUSSION AND CONCLUSIONS Chapter 6 112 6.1. MAIN OUTCOMES OF THE THESIS The concern about OMPs emissions into the environment has been increasing in the last decades due to the growing consumption of pharmaceuticals, personal care products, industrial chemicals and many other organic substances. These compounds pose negative effects to both the ecosystem and human health, making it necessary to reduce their concentrations in WWTPs effluents, which are the main source of OMPs discharge to the environment (Margot et al., 2015). Yet, the elimination efficiencies of many OMPs in WWTPs are highly variable or very low. For such reason, research has traditionally focused mostly on finding the most appropriate operational and environmental conditions to improve OMPs removals in WWTPs (Alvarino et al., 2018a). Nonetheless, specific changes in the operational parameters have led to different outcomes in terms of elimination due to their variable effect on the microbial populations and metabolic pathways (Ngo et al., 2020). Furthermore, certain environmental conditions may favor the elimination of some OMPs but, at the same time, hinder that of others (Alvarino et al., 2018b; Harb et al., 2019). Therefore, to establish long-lasting solutions to reduce OMPs concentrations globally in WWTPs effluents, it is necessary to have a deep understanding of the biotransformation mechanisms and to elucidate the bacterial and enzymatic key players. Most WWTPs consist of an activated sludge system comprising both autotrophic nitrifying and heterotrophic activities (Metcalf & Eddy, 2014). To date, multiple studies are dealing with the role of nitrifiers during OMPs biotransformation, but there is scarce information about the heterotrophic biotransformation mechanisms due to their huge and complex metabolic network (Fischer and Majewsky, 2014). However, previous works have shown their potential to contribute to OMPs removal and reduce the accumulation of TPs formed by nitrifying bacteria (Tran et al., 2013; Wu et al., 2020). Besides, the heterotrophic activity is gaining interest in the development of innovative and more energy-efficient WWTPs (Liu et al., 2020). Thus, this thesis tries to provide insight and increase the knowledge into the OMPs biotransformation mechanisms in aerobic heterotrophic conditions, aiming to solve the aforementioned questions (Chapter 1). In the following sections of this chapter, the main outcomes of this thesis are integrated and jointly discussed. General discussion and conclusions 113 6.1.1. Are aerobic heterotrophs effective to biotransform OMPs? The contribution of aerobic heterotrophic microorganisms to the removal of OMPs in activated sludge systems has generally been underestimated. The results shown in Chapters 3 and 4 prove that heterotrophs can biotransform a range of OMPs with broad physico-chemical properties to extents similar to those reported in the literature for activated sludge and purely nitrifying systems. In fact, for some compounds, such as SMX, DCF and TCS (Chapter 4), the heterotrophic biotransformation seems to be better than that of nitrifiers. These results suggest that the A-stages of novel WWTPs, working at high OLRs and short SRTs and operating primarily under heterotrophic conditions (Jimenez et al., 2015), could be able to efficiently and extensively biotransform OMPs before they reach the B-stage, which is proposed to be operated by an anammox consortium (Liu et al., 2020). Additionally, these findings question the need of promoting the activity of autotrophic nitrifiers in WWTPs, which, compared to heterotrophs, are slower growers and possess a less efficient metabolism and a lower diversity of organisms (Holtmann and Sell, 2002; K. Kim et al., 2020). However, the combination of both microbial populations could still be required to reduce the accumulation of some TPs, that may be formed by specific microorganisms but require others for their complete mineralization (Khunjar et al., 2011; Wu et al., 2020). So far, aerobic heterotrophs are the only microbial population that has proven their capacity to metabolize OMPs (Tran et al., 2013), raising the question about the relevance of metabolism and cometabolism in WWTPs (Alvarino et al., 2018a; Fischer and Majewsky, 2014). However, most previous studies have been performed with pure or enriched cultures and/or using OMPs concentrations considerably higher than those typically reported in WWTPs influents, making it difficult to extrapolate lab results to real-scale plants. In this thesis, seeking to match WWTPs conditions as much as possible, continuous reactors were operated long-term using activated sludge as the inoculum and spiking OMPs in the low μg L-1 range. The results proved that cometabolism was the preferential biotransformation mechanism under aerobic heterotrophic conditions and that the fate of OMPs is linked to that of primary substrates (Chapter 3). Accordingly, the enzymes with low substrate specificity present in the reactor are responsible for OMPs biotransformation thanks to similarities in the chemical structures of the OMPs and the enzyme’s natural substrates. Chapter 6 114 The biotransformation kinetic constant (kbiol) provides information about the biodegradability of OMPs (Han Tran et al., 2017) and, for the same compound, can greatly vary between WWTPs (Helbling et al., 2015; van Bergen et al., 2021). Such variability is generally attributed to changes in operational and environmental conditions that lead to variations in the microbiome and, consequently, in its biotransformation capacity. In Chapter 3, it is shown that increases in the intensity of the heterotrophic metabolic activity also lead to changes in kbiol, improving OMPs biotransformation. Thus, kbiol is not only dependent on the physico-chemical properties of the OMPs and the conditions applied to the WWTPs, but also on the metabolic activity of the biomass. This finding implies that the management of the organic substrates and the OLR can be efficient tools to control the microbial activity and, subsequently, OMPs biotransformation. Nonetheless, the changes observed in kbiol proved to be compoundspecific, suggesting that variations in the metabolic activity carry over unevenly across the biotransformation rates of the OMPs. Besides, if the HRT applied is high enough, improved biotransformation rates may not translate into better biotransformation extents. Overall, aerobic heterotrophic microorganisms can extensively biotransform OMPs through cometabolism and influence the biotransformation rate through changes in kbiol caused by variations in the metabolic activity. 6.1.2. Transformation products as a tool to identify key enzymatic activities Biotransformation reactions occur thanks to the versatile catalytic activity of the enzymes involved in the biodegradation of primary substrates. Identifying such key enzymes is a complex task that is still in its early stages (Achermann et al., 2020; Krah et al., 2016; Wicker et al., 2016; Zhou et al., 2015) and, in aerobic heterotrophic conditions, there are no conclusive studies about the enzymatic activities involved in OMPs biotransformation. Aiming to fill in that gap, several experiments with heterotrophic sludge were carried out in this thesis through two different methodologies : (i) TPs and (ii) -omic analyses. TPs analysis has been traditionally challenging due to the low OMPs concentrations found in environmental samples and the complex matrices where they are present (L. Zhang et al., 2013). However, TPs studies aiming at understanding their ecotoxicity and risk to the environment and the biotransformation limitations of OMPs have become more frequent in recent times thanks to major advances in analytical techniques (Fenner et al., General discussion and conclusions 115 2021; Gulde et al., 2016). Additionally, TPs structure analysis can be used as a prediction tool to determine the main enzymatic activities involved in OMPs biotransformation (Chapter 1). If primary substrates and TPs are known, it is possible to perform educated guesses about the enzymes carrying out the process. Applying this methodology, in this thesis (Chapter 4) it was possible to identify mono- and dioxygenases, dehydrogenases, hydrolases and transferases as some of the main enzymatic activities likely responsible for OMPs biotransformation under heterotrophic conditions. The reactions catalyzed by these enzymes involve oxidation, hydrolysis and conjugation steps. Oxidation, which is frequently an initial biotransformation step for many OMPs in activated sludge systems, was the main biotransformation reaction observed. Among oxidative steps, hydroxylation was the most predominant, being detected for multiple compounds, although oxidation through deamination, demethylation and dehydrogenation reactions was also determined. Interestingly, for some OMPs, different TPs were identified, suggesting that sequences of transformation reactions may take place at different functional groups. Besides, some of the TPs detected have been described as active metabolites with long-half lives, highlighting the need to elucidate their fate in WWTPs and perform ecotoxicity analyses. Identifying TPs can be a challenging task because OMPs can undergo a huge array of biotransformation reactions, leading to a tremendously large number of possible TPs. Therefore, the application of modeling tools, such as the EAWAG-PPS (EAWAGBBD/PPS) (Chapter 4), which are based on biotransformation rules supported by multiple literature studies, have proven to be of great aid to create lists of suspect TPs. Their use, along with lab experiments that serve as a confirmation method, is recommended to build pathway knowledge and predict the formation and toxicity of TPs. Overall, this thesis proved that TPs analysis is a useful tool to unravel relevant enzymes taking part in biotransformation reactions. However, this methodology alone does not allow to obtain direct and complete confirmation of the responsible enzymatic activities. Thus, it is advisable to combine the TPs results with those from other analytical methods, such as in vitro assays, or omic techniques, as discussed in the following section. 6.1.3. Omic techniques as a tool to identify key microbial and enzymatic players Metagenomic, metatranscriptomic and metaproteomic techniques are evolving rapidly and meta-omics association studies have started to be used to identify key Chapter 6 116 microbial and enzymatic players involved in OMPs biotransformation pathways (Achermann et al., 2020; Du et al., 2017; Fenner et al., 2021). Their potential is based on the fact that introducing OMPs in a specific environment affects the structure of the microbiome, leading to variations in the expression of genes and gene products and allowing to find biomarkers that help understand and predict the fate of OMPs in WWTPs. However, the application of this approach still raises some doubts; mainly because (i) many false positives leading to non-causal correlations can be easily generated and (ii) it is still necessary to confirm that the overexpression of gene products happens even when OMPs are added at trace level concentrations (Fenner et al., 2021). In Chapter 5, the characteristic trends followed by five bacterial families belonging to the phylum Proteobacteria and the key role of Corynebacterium during the heterotrophic biotransformation of SMX could be determined thanks to variations in their abundance measured through the application of DNA metabarcoding and metaproteomic techniques. Besides, a differential expression of five enzymes involved in central metabolic functions of Corynebacterium was also observed, suggesting its relevance during SMX biotransformation. Interestingly, these findings point towards the importance of the TCA cycle enzymes during OMPs biotransformation, which has not been reported before. It is unlikely that these enzymes participate in the initial biotransformation steps given the usual size and structure of OMPs, but they could be involved in the conversion of smaller TPs formed after several biotransformation steps, such as those belonging to the SMX-pterin conjugation pathway. Therefore, results highlight that omic techniques are useful to unravel biotransformation pathways and identify relevant microorganisms and enzymatic activities. Nonetheless, some enzymes that should have been present in the sludge could not be detected through metaproteomics (Chapter 5), possibly due to their low abundance compared to that of housekeeping or central metabolism proteins, which shows that technical developments to improve enzyme identification and quantification are still required. Additionally, in Chapter 5,Corynebacterium abundances differed between the results from DNA metabarcoding and metaproteomic analyses. Thus, omic techniques representing expressed gene products at a given time, such as metatranscriptomics or metaproteomics, may be more suitable to find causal links between OMPs biotransformation and microbiological data. General discussion and conclusions 117 6.2. MAIN GAPS AND FUTURE PERSPECTIVES There are several research areas and knowledge gaps that need to be addressed to better understand biological processes in WWTPs and improve OMPs removal. Cometabolism has been identified as the main mechanism driving OMPs biotransformation; however, further research is required to determine where the boundary between metabolism and cometabolism lies (Nsenga Kumwimba and Meng, 2019; Tran et al., 2013). To this end, experiments with increasing OMPs doses, along with -omics and TPs analyses, could be carried out. These experiments would also allow determining the threshold where inhibitory events appear and provide information about changes in biotransformation routes at higher OMPs concentrations (Jia et al., 2020; Wegner et al., 2015), which could be particularly useful for highly concentrated streams, such as hospital or industrial wastewaters. Moreover, the contribution of secondary biological activities of the sludge during the cometabolic biotransformation of OMPs remains unclear (Gonzalez-Gil et al., 2021). The more recent use of radiolabeling-based approaches, which consist of the addition of a labeled substrate to the experimental environment along with an analysis of label incorporation into TPs, biomarkers, enzymes or cells, has been pointed out as a promising technique to solve these questions. Yet, its application with OMPs present at low levels remains challenging (Falås et al., 2018; Fenner et al., 2021). Insight on the role and origin of the enzymes responsible for OMPs biotransformation, as well as on the conditions that increase their expression, is necessary to implement operational strategies in WWTPs that enhance their abundances and activities (Fischer and Majewsky, 2014; Krah et al., 2016). Such knowledge could eventually allow to select enzymes as a bioremediation technique to target specific compounds, particularly for the treatment of well-defined waste streams. It would offer a more direct, controlled and defined alternative than conventional WWTPs, which comprise a huge consortium of microorganisms and enzymes and require the control of cell growth, substrate transport to the cell and sorption, among others (Stadlmair et al., 2018). Nonetheless, the application of this technique in real systems is still a long way off for several reasons: (i) the isolation of target enzymes from WWTPs has not been achieved yet, (ii) the purchase of enzymes is too costly and presents limitations of largescale production and (iii) setting up cooperation between multiple enzymes in WWTPs Chapter 6 118 and dealing with enzymatic inactivation, stability and specificity is still a complex task (Feng et al., 2021; Langbehn et al., 2021; Stadlmair et al., 2018). To date, although some approaches have been developed and applied to identify the enzymes involved in OMPs biotransformation, they still present multiple limitations. For instance, enzymatic in vitro assays with cell-free lysates require further research on the extraction procedure to preserve the indigenous enzymatic activities present in the sludge, including the application of pretreatments and buffer additives (Krah et al., 2016). Moreover, it is necessary to improve the understanding of the circumstances in which in vitro assays are fully comparable to real biological processes and to evaluate the adequacy of applying cofactors or enzymatic inhibitors, that may boost or hinder specific enzymes and help to confirm their activity towards OMPs (Gonzalez-Gil et al., 2019b). To facilitate the link between enzymes and biotransformation pathways, the detection, quantification and assessment of TPs require the continuous improvement of modeling tools and analytical techniques that allow sensitive and untargeted TPs characterization in short time frames even when reference standards are not available (Nguyen et al., 2021). In the past decades, high-resolution mass spectrometry has resulted in notorious advances to determine low OMPs and TPs concentrations even in complex mixtures. More recently, compound-specific isotope analysis is being developed, although its application to environmental OMPs concentrations remains a challenge. This technique can potentially help to determine the biotransformation extent and reaction mechanisms of OMPs and elucidate different origins of the same TPs thanks to the usual enzymatic preference for molecules with light isotopes, which leads to changes in the isotopic ratios of the biotransformed compounds (Fenner et al., 2021). Finally, omics techniques need to reduce their costs to broaden their accessibility and their results must be supported with powerful statistical tools to facilitate the search of causal relationships between enzymes and OMPs biotransformation (Achermann et al., 2020; David R. Johnson et al., 2015b). As a result of the abovementioned limitations, understanding the influence of environmental and process conditions and confidently obtaining mechanistic insights about the biotransformation of OMPs in WWTPs by directly and exclusively applying any of the well-established approaches is still not possible. However, a combination of all methodological approaches is the best way to overcome some of the individual drawbacks and formulate sound hypotheses about the microorganisms and enzymes participating in OMPs biotransformation, in line with previous studies (Achermann et al., General discussion and conclusions 119 2018b, 2020; Zhao et al., 2020) and the results of this thesis (Chapter 5). Thus, to maximize OMPs removal is essential an interdisciplinary contribution from the scientific community, covering the fields of environmental engineering, analytical chemistry, molecular biology and data sciences, as well as validating laboratory findings in full-scale mixed cultures with a wide range of OMPs present at environmentally relevant concentrations. References 126 Criddle, C.S., 1993. The kinetics of cometabolism. Biotechnol. Bioeng. 41, 1048–1056. https://doi.org/10.1002/bit.260411107 Cydzik-Kwiatkowska, A., Zielińska, M., Bernat, K., Bułkowska, K., Wojnowska-Baryła, I., 2020. Insights into mechanisms of bisphenol A biodegradation in aerobic granular sludge. Bioresour. Technol. 315, 123806. https://doi.org/10.1016/j.biortech.2020.123806 Davids, M., Gudra, D., Radovica-Spalvina, I., Fridmanis, D., Bartkevics, V., Muter, O., 2017. The effects of ibuprofen on activated sludge: Shift in bacterial community structure and resistance to ciprofloxacin. J. Hazard. Mater. 340, 291–299. https://doi.org/10.1016/j.jhazmat.2017.06.065 Di Marcantonio, C., Chiavola, A., Bains, A., Singhal, N., 2020. Effect of oxic/anoxic conditions on the removal of organic micropollutants in the activated sludge process. Environ. Technol. Innov. 20, 101161. https://doi.org/10.1016/j.eti.2020.101161 Ding, T., Li, W., Cai, M., Jia, X., Yang, M., Yang, B., Li, J., 2020. Algal toxicity, accumulation and metabolic pathways of galaxolide. J. Hazard. Mater. 384, 121360. https://doi.org/10.1016/j.jhazmat.2019.121360 Du, Z., Chen, Y., Li, X., 2017. Quantitative proteomic analyses of the microbial degradation of estrone under various background nitrogen and carbon conditions. Water Res. 123, 361–368. https://doi.org/10.1016/j.watres.2017.06.070 Dueholm, M.S., Nierychlo, M., Andersen, K.S., Rudkjøbing, V., Knutsson, S., Consortium, the M.G., Albertsen, M., Nielsen, P.H., 2021. MiDAS 4: A global catalogue of full-length 16S rRNA gene sequences and taxonomy for studies of bacterial communities in wastewater treatment plants. BioRxiv 2021.07.06.451231. https://doi.org/10.1101/2021.07.06.451231 EAWAG-BBD Pathway Prediction System (available at http://eawagbbd.ethz.ch/predict/), 2021. Ejhed, H., Fång, J., Hansen, K., Graae, L., Rahmberg, M., Magnér, J., Dorgeloh, E., Plaza, G., 2018. The effect of hydraulic retention time in onsite wastewater treatment and removal of pharmaceuticals, hormones and phenolic utility substances. Sci. Total Environ. 618, 250–261. https://doi.org/10.1016/j.scitotenv.2017.11.011 European Commission, 2020. Comission Implementing Decision (EU) 2020/1161. Off. J. Eur. Union L 257, 32–35. Falås, P., Jewell, K.S., Hermes, N., Wick, A., Ternes, T.A., Joss, A., Nielsen, J.L., 2018. Transformation, CO2 formation and uptake of four organic micropollutants by References 127 carrier-attached microorganisms. Water Res. 141, 405–416. https://doi.org/10.1016/j.watres.2018.03.040 Falås, P., Longrée, P., La Cour Jansen, J., Siegrist, H., Hollender, J., Joss, A., 2013. Micropollutant removal by attached and suspended growth in a hybrid biofilmactivated sludge process. Water Res. 47, 4498–4506. https://doi.org/10.1016/j.watres.2013.05.010 Falås, P., Wick, A., Castronovo, S., Habermacher, J., Ternes, T.A., Joss, A., 2016. Tracing the limits of organic micropollutant removal in biological wastewater treatment. Water Res. 95, 240–249. https://doi.org/10.1016/j.watres.2016.03.009 Fedeles, B.I., Singh, V., Delaney, J.C., Li, D., Essigmann, J.M., 2015. The AlkB family of Fe(II)/α-ketoglutarate-dependent dioxygenases: Repairing nucleic acid alkylation damage and beyond. J. Biol. Chem. https://doi.org/10.1074/jbc.R115.656462 Feng, S., Hao Ngo, H., Guo, W., Woong Chang, S., Duc Nguyen, D., Cheng, D., Varjani, S., Lei, Z., Liu, Y., 2021. Roles and applications of enzymes for resistant pollutants removal in wastewater treatment. Bioresour. Technol. https://doi.org/10.1016/j.biortech.2021.125278 Fenner, K., Elsner, M., Lueders, T., McLachlan, M.S., Wackett, L.P., Zimmermann, M., Drewes, J.E., 2021. Methodological Advances to Study Contaminant Biotransformation: New Prospects for Understanding and Reducing Environmental Persistence? Environ. Sci. Technol. https://doi.org/10.1021/acsestwater.1c00025 Fernandez-Fontaina, E., Carballa, M., Omil, F., Lema, J.M., 2014. Modelling cometabolic biotransformation of organic micropollutants in nitrifying reactors. Water Res. 65, 371–383. https://doi.org/10.1016/j.watres.2014.07.048 Fernandez-Fontaina, E., Gomes, I.B., Aga, D.S., Omil, F., Lema, J.M., Carballa, M., 2016. Biotransformation of pharmaceuticals under nitrification, nitratation and heterotrophic conditions. Sci. Total Environ. 541, 1439–1447. https://doi.org/10.1016/j.scitotenv.2015.10.010 Fernandez-Fontaina, E., Omil, F., Lema, J.M., Carballa, M., 2012. Influence of nitrifying conditions on the biodegradation and sorption of emerging micropollutants. Water Res. 46, 5434–5444. https://doi.org/10.1016/j.watres.2012.07.037 Fernandez-Fontaina, E., Pinho, I., Carballa, M., Omil, F., Lema, J.M., 2013. Biodegradation kinetic constants and sorption coefficients of micropollutants in membrane bioreactors. Biodegradation 24, 165–177. https://doi.org/10.1007/s10532-012-9568-3 References 128 Fischer, K., Majewsky, M., 2014. Cometabolic degradation of organic wastewater micropollutants by activated sludge and sludge-inherent microorganisms. Appl. Microbiol. Biotechnol. 98, 6583–6597. https://doi.org/10.1007/s00253-014-5826-0 Gallardo-Altamirano, M.J., Maza-Márquez, P., Montemurro, N., Rodelas, B., Osorio, F., Pozo, C., 2019. Linking microbial diversity and population dynamics to the removal efficiency of pharmaceutically active compounds (PhACs) in an anaerobic/anoxic/aerobic (A2O) system. Chemosphere 233, 828–842. https://doi.org/10.1016/j.chemosphere.2019.06.017 Ganem, B., 1995. The hydrolysis of amines to alcohols: clues from chrorismate processing enzymes. Tetrahedron Lett. 35, 815–818. Gardner, M., Jones, V., Comber, S., Scrimshaw, M.D., Coello-Garcia, T., Cartmell, E., Lester, J., Ellor, B., 2013. Performance of UK wastewater treatment works with respect to trace contaminants. Sci. Total Environ. 456–457, 359–369. https://doi.org/10.1016/j.scitotenv.2013.03.088 Gaulke, L.S., Strand, S.E., Kalhorn, T.F., Stensel, H.D., 2008. 17Α-Ethinylestradiol Transformation Via Abiotic Nitration in the Presence of Ammonia Oxidizing Bacteria. Environ. Sci. Technol. 42, 7622–7627. https://doi.org/10.1021/es801503u Gonzalez-Gil, L., Carballa, M., Corvini, P.F.X., Lema, J.M., 2019a. Reversibility of enzymatic reactions might limit biotransformation of organic micropollutants. Sci. Total Environ. 665, 574–578. https://doi.org/10.1016/j.scitotenv.2019.02.143 Gonzalez-Gil, L., Carballa, M., Lema, J.M., 2017. Cometabolic Enzymatic Transformation of Organic Micropollutants under Methanogenic Conditions. Environ. Sci. Technol. 51, 2963–2971. https://doi.org/10.1021/acs.est.6b05549 Gonzalez-Gil, L., Fernandez-Fontaina, E., Singh, R.R., Lema, J.M., Carballa, M., Aga, D.S., 2021. Feeding composition and sludge retention time both affect ( co- )metabolic biotransformation of pharmaceutical compounds in activated sludge systems. J. Environ. Chem. Eng. Gonzalez-Gil, L., Krah, D., Ghattas, A.K., Carballa, M., Wick, A., Helmholz, L., Lema, J.M., Ternes, T.A., 2019b. Biotransformation of organic micropollutants by anaerobic sludge enzymes. Water Res. 152, 202–214. https://doi.org/10.1016/j.watres.2018.12.064 Gonzalez-Gil, L., Mauricio-Iglesias, M., Carballa, M., Lema, J.M., 2018a. Why are organic micropollutants not fully biotransformed? A mechanistic modelling approach to anaerobic systems. Water Res. 142, 115–128. References 129 https://doi.org/https://doi.org/10.1016/j.watres.2018.05.032 Gonzalez-Gil, L., Mauricio-Iglesias, M., Serrano, D., Lema, J.M., Carballa, M., 2018b. Role of methanogenesis on the biotransformation of organic micropollutants during anaerobic digestion. Sci. Total Environ. 622–623, 459–466. https://doi.org/10.1016/j.scitotenv.2017.12.004 Gulde, R., Anliker, S., Kohler, H.P.E., Fenner, K., 2018. Ion Trapping of Amines in Protozoa: A Novel Removal Mechanism for Micropollutants in Activated Sludge. Environ. Sci. Technol. 52, 52–60. https://doi.org/10.1021/acs.est.7b03556 Gulde, R., Helbling, D.E., Scheidegger, A., Fenner, K., 2014. PH-dependent biotransformation of ionizable organic micropollutants in activated sludge. Environ. Sci. Technol. 48, 13760–13768. https://doi.org/10.1021/es5037139 Gulde, R., Meier, U., Schymanski, E.L., Kohler, H.P.E., Helbling, D.E., Derrer, S., Rentsch, D., Fenner, K., 2016. Systematic Exploration of Biotransformation Reactions of Amine-Containing Micropollutants in Activated Sludge. Environ. Sci. Technol. 50, 2908–2920. https://doi.org/10.1021/acs.est.5b05186 Han, P., Yu, Y., Zhou, L., Tian, Z., Li, Z., Hou, L., Liu, M., Wu, Q., Wagner, M., Men, Y., 2019. Specific Micropollutant Biotransformation Pattern by the Comammox Bacterium Nitrospira inopinata. Environ. Sci. Technol. 53, 8695–8705. https://doi.org/10.1021/acs.est.9b01037 Han Tran, N., Reinhard, M., Yew-Hoong Gin, K., 2017. Occurrence and fate of emerging contaminants in municipal wastewater treatment plants from different geographical regions- A review. Water Res. https://doi.org/10.1016/j.watres.2017.12.029 Harb, M., Lou, E., Smith, A.L., Stadler, L.B., 2019. Perspectives on the fate of micropollutants in mainstream anaerobic wastewater treatment. Curr. Opin. Biotechnol. 57, 94–100. https://doi.org/10.1016/j.copbio.2019.02.022 Harb, M., Wei, C.H., Wang, N., Amy, G., Hong, P.Y., 2016. Organic micropollutants in aerobic and anaerobic membrane bioreactors: Changes in microbial communities and gene expression. Bioresour. Technol. 218, 882–891. https://doi.org/10.1016/j.biortech.2016.07.036 Helbling, D.E., Hollender, J., Kohler, H.E., 2010a. Structure-Based Interpretation of Biotransformation Pathways of Amide-Containing Compounds in Sludge-Seeded Bioreactors. Environ. Sci. Technol. 44, 6628–6635. https://doi.org/10.1021/es101035b Helbling, D.E., Hollender, J., Kohler, H.P.E., Singer, H., Fenner, K., 2010b. High- References 130 throughput identification of microbial transformation products of organic micropollutants. Environ. Sci. Technol. 44, 6621–6627. https://doi.org/10.1021/es100970m Helbling, D.E., Johnson, D.R., Honti, M., Fenner, K., 2012. Micropollutant biotransformation kinetics associate with WWTP process parameters and microbial community characteristics. Environ. Sci. Technol. 46, 10579–10588. https://doi.org/10.1021/es3019012 Helbling, D.E., Johnson, D.R., Lee, T.K., Scheidegger, A., Fenner, K., 2015. A framework for establishing predictive relationships between specific bacterial 16S rRNA sequence abundances and biotransformation rates. Water Res. 70, 471–484. https://doi.org/10.1016/j.watres.2014.12.013 Holtmann, D., Sell, D., 2002. Detection of the microbial activity of aerobic heterotrophic, anoxic heterotrophic and aerobic autotrophic activated sludge organisms with an electrochemical sensor. Biotechnol. Lett. 24, 1313–1318. https://doi.org/10.1023/A:1019871912731 Jewell, K.S., Castronovo, S., Wick, A., Falås, P., Joss, A., Ternes, T.A., 2016a. New insights into the transformation of trimethoprim during biological wastewater treatment. Water Res. 88, 550–557. https://doi.org/10.1016/j.watres.2015.10.026 Jewell, K.S., Falås, P., Wick, A., Joss, A., Ternes, T.A., 2016b. Transformation of diclofenac in hybrid biofilm–activated sludge processes. Water Res. 105, 559–567. https://doi.org/10.1016/j.watres.2016.08.002 Jia, Y., Yin, L., Khanal, S.K., Zhang, H., Oberoi, A.S., Lu, H., 2020. Biotransformation of ibuprofen in biological sludge systems: Investigation of performance and mechanisms. Water Res. 170, 115303. https://doi.org/10.1016/j.watres.2019.115303 Jiang, C., Geng, J., Hu, H., Ma, H., Gao, X., Ren, H., 2017. Impact of selected nonsteroidal anti-inflammatory pharmaceuticals on microbial community assembly and activity in sequencing batch reactors. PLoS One 12, e0179236. https://doi.org/10.1371/journal.pone.0179236 Jimenez, J., Miller, M., Bott, C., Murthy, S., De Clippeleir, H., Wett, B., 2015. High-rate activated sludge system for carbon management - Evaluation of crucial process mechanisms and design parameters. Water Res. 87, 476–482. https://doi.org/10.1016/j.watres.2015.07.032 Johnson, David R., Helbling, D.E., Lee, T.K., Park, J., Fenner, K., Kohler, H.P.E., Ackermann, M., 2015a. Association of biodiversity with the rates of micropollutant References 131 biotransformations among full-scale wastewater treatment plant communities. Appl. Environ. Microbiol. 81, 666–675. https://doi.org/10.1128/AEM.03286-14 Johnson, David R., Helbling, D.E., Men, Y., Fenner, K., 2015b. Can meta-omics help to establish causality between contaminant biotransformations and genes or gene products? Environ. Sci. Water Res. Technol. https://doi.org/10.1039/c5ew00016e Kaiser, E., Prasse, C., Wagner, M., Bröder, K., Ternes, T.A., 2014. Transformation of oxcarbazepine and human metabolites of carbamazepine and oxcarbazepine in wastewater treatment and sand filters. Environ. Sci. Technol. 48, 10208–10216. https://doi.org/10.1021/es5024493 Kennes-Veiga, D.M., Gónzalez-Gil, L., Carballa, M., Lema, J.M., 2021a. Enzymatic cometabolic biotransformation of organic micropollutants in wastewater treatment plants: A review. Bioresour. Technol. 126291. https://doi.org/10.1016/j.biortech.2021.126291 Kennes-Veiga, D.M., Gónzalez-Gil, L., Carballa, M., Lema, J.M., 2020. The organic loading rate affects organic micropollutants’ cometabolic biotransformation kinetics under heterotrophic conditions in activated sludge. Water Res. 189, 116587. https://doi.org/10.1016/j.watres.2020.116587 Kennes-Veiga, D.M., Vogler, B., Fenner, K., Carballa, M., Lema, J.M., 2021b. Heterotrophic enzymatic biotransformations of organic micropollutants in activated sludge. Sci. Total Environ. 780, 146564. https://doi.org/10.1016/j.scitotenv.2021.146564 Kern, S., Baumgartner, R., Helbling, D.E., Hollender, J., Singer, H., Loos, M.J., Schwarzenbach, R.P., Fenner, K., 2010. A tiered procedure for assessing the formation of biotransformation products of pharmaceuticals and biocides during activated sludge treatment. J. Environ. Monit. 12, 2100–2111. https://doi.org/10.1039/c0em00238k Khunjar, W.O., MacKintosh, S.A., Skotnicka-Pitak, J., Baik, S., Aga, D.S., Love, N.G., 2011. Elucidating the relative roles of ammonia oxidizing and heterotrophic bacteria during the biotransformation of 17α-ethinylestradiol and trimethoprim. Environ. Sci. Technol. 45, 3605–3612. https://doi.org/10.1021/es1037035 Kim, J.Y., Ryu, K., Kim, E.J., Choe, W.S., Cha, G.C., Yoo, I.K., 2007. Degradation of bisphenol A and nonylphenol by nitrifying activated sludge. Process Biochem. 42, 1470–1474. https://doi.org/10.1016/j.procbio.2007.06.010 Kim, K., Hur, J.W., Kim, S., Jung, J.Y., Han, H.S., 2020. Biological wastewater References 132 treatment: Comparison of heterotrophs (BFT) with autotrophs (ABFT) in aquaculture systems. Bioresour. Technol. 296. https://doi.org/10.1016/j.biortech.2019.122293 Kim, M.H., Fan, C., Pan, S.-Y., Lee, I., Lin, Y., Kim, H., 2020. Kinetics of competitive cometabolism under aerobic conditions. Water-Energy Nexus 3, 62–70. https://doi.org/10.1016/j.wen.2020.04.001 King, R.B., Sheldon, J.K., Long, G.M., 1997. Practical Environmental Bioremediation: The Field Guide, 2nd ed. Kolvenbach, B.A., Helbling, D.E., Kohler, H.P.E., Corvini, P.F.X., 2014. Emerging chemicals and the evolution of biodegradation capacities and pathways in bacteria. Curr. Opin. Biotechnol. https://doi.org/10.1016/j.copbio.2013.08.017 Kora, E., Theodorelou, D., Gatidou, G., Fountoulakis, M.S., Stasinakis, A.S., 2020. Removal of polar micropollutants from domestic wastewater using a methanogenic –aerobic moving bed biofilm reactor system. Chem. Eng. J. 382, 122983. https://doi.org/10.1016/j.cej.2019.122983 Kosjek, T., Perko, S., Zupanc, M., ZanoŠki Hren, M., Landeka Dragičević, T., Žigon, D., Kompare, B., Heath, E., 2012. Environmental occurrence, fate and transformation of benzodiazepines in water treatment. Water Res. 46, 355–368. https://doi.org/10.1016/j.watres.2011.10.056 Krah, D., Ghattas, A.K., Wick, A., Bröder, K., Ternes, T.A., 2016. Micropollutant degradation via extracted native enzymes from activated sludge. Water Res. 95, 348–360. https://doi.org/10.1016/j.watres.2016.03.037 Kruglova, A., Kråkström, M., Riska, M., Mikola, A., Rantanen, P., Vahala, R., Kronberg, L., 2016. Comparative study of emerging micropollutants removal by aerobic activated sludge of large laboratory-scale membrane bioreactors and sequencing batch reactors under low-temperature conditions. Bioresour. Technol. 214, 81–88. https://doi.org/10.1016/j.biortech.2016.04.037 Lacerda, C.M.R., Reardon, K.F., 2009. Environmental proteomics: applications of proteome profiling in environmental microbiology and biotechnology. Brief. Funct. Genomics 8, 75–87. https://doi.org/10.1093/bfgp/elp005 Langbehn, R.K., Michels, C., Soares, H.M., 2021. Antibiotics in wastewater: From its occurrence to the biological removal by environmentally conscious technologies. Environ. Pollut. 275, 116603. https://doi.org/10.1016/j.envpol.2021.116603 Larcher, S., Yargeau, V., 2011. Biodegradation of sulfamethoxazole by individual and References 133 mixed bacteria. Appl. Microbiol. Biotechnol. 91, 211–218. https://doi.org/10.1007/s00253-011-3257-8 Lee, D.G., Cho, K.C., Chu, K.H., 2015. Removal of triclosan in nitrifying activated sludge: Effects of ammonia amendment and bioaugmentation. Chemosphere 125, 9– 15. https://doi.org/10.1016/j.chemosphere.2014.12.085 Lema, J.M., Suarez, S., 2017. Innovative Wastewater Treatment & Resource Recovery Technologies: Impacts on Energy, Economy and Environment, Water Intelligence Online. IWA Publishing. https://doi.org/10.2166/9781780407876 Lemmer, H., Nitschke, L., 1994. Vitamin content of four sludge fractions in the activated sludge wastewater treatment process. Water Res. 28, 737–739. https://doi.org/10.1016/0043-1354(94)90155-4 Li, B., Qiu, Y., Shi, H., Yin, H., 2016. The importance of lag time extension in determining bacterial resistance to antibiotics. Analyst 141, 3059–3067. https://doi.org/10.1039/c5an02649k Li, F., Yuasa, A., Obara, A., Mathews, A.P., 2005. Aerobic batch degradation of 17-β estradiol (E2) by activated sludge: Effects of spiking E2 concentrations, MLVSS and temperatures. Water Res. 39, 2065–2075. https://doi.org/10.1016/j.watres.2005.02.009 Li, X., Xu, Q.M., Cheng, J.S., Yuan, Y.J., 2016. Improving the bioremoval of sulfamethoxazole and alleviating cytotoxicity of its biotransformation by laccase producing system under coculture of Pycnoporus sanguineus and Alcaligenes faecalis. Bioresour. Technol. 220, 333–340. https://doi.org/10.1016/J.BIORTECH.2016.08.088 Lin, S., Rong, K., Lamichhane, K.M., Babcock, R.W., Kirs, M., Cooney, M.J., 2020. Anaerobic-aerobic biofilm-based digestion of chemical contaminants of emerging concern (CEC) and pathogen indicator organisms in synthetic wastewater. Bioresour. Technol. 299. https://doi.org/10.1016/j.biortech.2019.122554 Liu, G., Wang, J., 2015. Modeling effects of DO and SRT on activated sludge decay and production. Water Res. 80, 169–178. https://doi.org/10.1016/j.watres.2015.04.042 Liu, R., Li, S., Tu, Y., Hao, X., 2021. Capabilities and mechanisms of microalgae on removing micropollutants from wastewater: A review. J. Environ. Manage. https://doi.org/10.1016/j.jenvman.2021.112149 Liu, Y., Gu, J., Zhang, M., 2020. A-B Processes: Towards Energy Self-sufficient Municipal Wastewater Treatment, IWA Publising. References 134 https://doi.org/10.2166/9781789060089 Ltd I C Consultants, 2001. Pollutants in urban wastewater and sewage sludge (European Comission report). Luk, S., Atayee, R.S., Ma, J.D., Best, B.M., 2014. Urinary diazepam metabolite distribution in a chronic pain population. J. Anal. Toxicol. 38, 135–142. https://doi.org/10.1093/jat/bku001 Luo, Y., Guo, W., Ngo, H.H., Nghiem, L.D., Hai, F.I., Zhang, J., Liang, S., Wang, X.C., 2014. A review on the occurrence of micropollutants in the aquatic environment and their fate and removal during wastewater treatment. Sci. Total Environ. 473–474, 619–641. https://doi.org/10.1016/j.scitotenv.2013.12.065 Ma, B., Zhang, K., Hendrie, C., Liang, C., Li, M., Doherty-Kirby, A., Lajoie, G., 2003. PEAKS: powerful software for peptide de novo sequencing by tandem mass spectrometry. Rapid Commun. Mass Spectrom. 17, 2337–2342. https://doi.org/https://doi.org/10.1002/rcm.1196 Majewsky, M., Gallé, T., Yargeau, V., Fischer, K., 2011. Active heterotrophic biomass and sludge retention time (SRT) as determining factors for biodegradation kinetics of pharmaceuticals in activated sludge. Bioresour. Technol. 102, 7415–7421. https://doi.org/10.1016/j.biortech.2011.05.032 Majewsky, M., Gallé, T., Zwank, L., Fischer, K., 2010. Influence of microbial activity on polar xenobiotic degradation in activated sludge systems. Water Sci. Technol. 62, 701–707. https://doi.org/10.2166/wst.2010.925 Majewsky, M., Glauner, T., Horn, H., 2015. Systematic suspect screening and identification of sulfonamide antibiotic transformation products in the aquatic environment. Anal. Bioanal. Chem. 407. https://doi.org/10.1007/s00216-015-8748- 5 Majewsky, M., Wagner, D., Delay, M., Bräse, S., Yargeau, V., Horn, H., 2014. Antibacterial activity of sulfamethoxazole transformation products (TPs): General relevance for sulfonamide TPs modified at the para position. Chem. Res. Toxicol. 27, 1821–1828. https://doi.org/10.1021/tx500267x Mandarić, L., Petrovic, M., Sabater, S., 2018. Transport, distribution and the fate of emerging contaminants in wastewater-receiving rivers under multiple stress conditions. Dr. Thesis. Catalan Institute for Water Research - Universitat de Girona. Margot, J., Lochmatter, S., Barry, D.A., Holliger, C., 2016. Role of ammonia-oxidizing bacteria in micropollutant removal from wastewater with aerobic granular sludge. References 135 Water Sci. Technol. 73, 564–575. https://doi.org/10.2166/wst.2015.514 Margot, J., Rossi, L., Barry, D.A., Holliger, C., 2015. A review of the fate of micropollutants in wastewater treatment plants. Wiley Interdiscip. Rev. Water 2, 457–487. https://doi.org/10.1002/wat2.1090 Martin, C., Moeder, M., Daniel, X., Krauss, G., 2007. Biotransformation of the Polycyclic Musks HHCB and AHTN and Metabolite Formation by Fungi Occurring in Freshwater Environments 41, 5395–5402. https://doi.org/https://doi.org/10.1021/es0711462 Martínez-Quintela, M., Arias, A., Alvarino, T., Suarez, S., Garrido, J.M., Omil, F., 2021. Cometabolic removal of organic micropollutants by enriched nitrite-dependent anaerobic methane oxidizing cultures. J. Hazard. Mater. 402. https://doi.org/10.1016/j.jhazmat.2020.123450 Men, Y., Achermann, S., Helbling, D.E., Johnson, D.R., Fenner, K., 2017. Relative contribution of ammonia oxidizing bacteria and other members of nitrifying activated sludge communities to micropollutant biotransformation. Water Res. 109, 217–226. https://doi.org/10.1016/j.watres.2016.11.048 Men, Y., Han, P., Helbling, D.E., Jehmlich, N., Herbold, C., Gulde, R., Onnis-Hayden, A., Gu, A.Z., Johnson, D.R., Wagner, M., Fenner, K., 2016. Biotransformation of Two Pharmaceuticals by the Ammonia-Oxidizing Archaeon Nitrososphaera gargensis. Environ. Sci. Technol. 50, 4682–4692. https://doi.org/10.1021/acs.est.5b06016 Messerschmidt, A., 2010. Copper metalloenzymes, in: Comprehensive Natural Products II: Chemistry and Biology. Elsevier Ltd, pp. 489–545. https://doi.org/10.1016/b978- 008045382-8.00180-5 Metcalf & Eddy, 2014. Wastewater Engineering: Treatment and Resource Recovery, 5th ed. McGraw-Hill, New York (USA). Moriya, Y., Shigemizu, D., Hattori, M., Tokimatsu, T., Kotera, M., Goto, S., Kanehisa, M., 2010. PathPred: An enzyme-catalyzed metabolic pathway prediction server. Nucleic Acids Res. 38. https://doi.org/10.1093/NAR/GKQ318 Moya-Llamas, M.J., Trapote, A., Prats, D., 2018. Removal of micropollutants from urban wastewater using a UASB reactor coupled to a MBR at different organic loading rates. Urban Water J. 15, 437–444. https://doi.org/10.1080/1573062X.2018.1508599 Mulla, S.I., Hu, A., Sun, Q., Li, J., Suanon, F., Ashfaq, M., Yu, C.P., 2018.