1 Introduction

Peroxidases (Per) are hemoproteins involved in the oxidation reactions with hydrogen peroxide (H2O2) and an electron donor compound. They belong to a large family of oxido-reductases and found in living organisms with size ranges between 35–100 kDalton [1]. Although Per could use a large variety of electron donor compounds, some use specific ones such as glutathione (glutathione peroxidases). Although Per have a major protection role against reactive oxygen species such as H2O2, they can lead to deleterious reactions involving the oxidation of endogenous substrates and xenobiotics leading to tissue damage, lipoprotein oxidation and carcinogenesis. They are considered antioxidant enzymes to keep H2O2 at safe levels to protect cells and tissues. Indeed, the estimated toxicity of H2O2 is 0.1 µg/L after 96 h at 17 °C in trout [2]. Given that between 1–2% of consumed O2 during respiration transforms into oxygen radicals (*OH) and to H2O2 by superoxide dismutase [3], the levels of H2O2 needs to be tightly regulated by antioxidant mechanisms involving catalase and Per. Moreover, the long-term of activity of Per could lead to the accumulation of oxidized compounds (from electron donors) such as oxidized vitamins/cofactors and xenobiotics leading to DNA adduct and DNA (8-oxoguanine) oxidation [4]. Indeed, reactive species generated by mitochondria or other sites in the cytoplasm, cause oxidative damage and initiate degradative processes involved in aging. Oxidative stress can modify lipids, DNA, RNA and proteins/enzymes, which require them to be removed by DNA repair enzymes, protein (protein degradation/turnover) and (damaged) unsaturated lipids (hydrolysis). Hence, the Per enzymes are two edged swords i.e., they can removes toxic levels of H2O2 at the expenses of electron donor products in cells (vitamins, NADH, guanosine, glutathione, amino acids etc.) leading to the accumulation of oxidatively damaged products on the long-term. In fish collected at rivers polluted by heavy metal, the fish accumulated zinc, chromium, nickel, cobalt and copper with a concomitant rise in Per activity in the liver, gills and muscles [5]. This suggests that these metals lead to H2O2 perhaps at the expense of antioxidant levels (i.e., reduced antoxidant capacity). From the wastewater management perspective, contaminants affecting this enzyme system could form the basis of toxicity initiation by the direct oxidation potential of H2O2 and the long-term loss of antioxidants in organisms. In this respect, changes in Per activity by wastewaters could serve as a first signal leading to toxicity in exposed organisms.

The Per assay was used to screen various industrial effluents and revealed inhibition in the reaction rates of H2O2 hydrolysis [6]. It was noteworthy that the Per inhibition potential was significantly associated to trout mortality where the toxic effluents produced the strongest inhibition at low effluent concentration. An interesting variation of this assay was to include DNA during the Per assay to detect interactions between DNA and the effluents. While DNA alone did not affect Per activity, the addition of DNA with the effluents reversed the inhibition suggesting interactions with the contaminants. Effluents showing DNA interactions with the Per reaction were genotoxic 70% of the cases using an SOS DNA repair test in bacteria [6]. A luminol-based Per reaction was also proposed as a mean for water quality assessments with single substances and pesticide formulations [7]. Preincubation of Per with herbicides, detergents/surfactant, phenol, metals (Hg, Co, Ni) had an inhibitory effect of Per activity. However, some other compounds could stimulate Per activity at low concentrations (an hormetic response), which was followed by decreased activity. Some insecticide formulation revealed a sustained increase in Per activity suggesting perhaps that other compounds present in these formulations (stabilizers, antioxidants etc.) could increase H2O2 degradation rates. The Per reaction offers a very rapid and efficient means to screen for potentially toxic mixtures such as industrial and municipal wastewaters. In the context of reducing the sacrifice of fish and other vertebrates in toxicity tests, new alternative methods (NAMs) are urgently needed for toxicity screening purposes and the Per assay was seemingly predictive of trout mortality with very few or no false negatives. These quick and rapid biosensors could also be used with machine learning (artificial intelligence) for forecasting impacts based on climate (rain falls), input of industrial and domestic sewage, wastewater treatment plants capacity and transit times [8]. At present, the dynamic understanding of the influence of population size, climate changes (rain forecasts) on the potential toxicity of various types of wastewater treatment are lacking where quick rapid assays coupled to wastewater treatments during various extremes climate events could bridge this gap.

The purpose of this study was therefore to use the Per and DNA-protection assays as a screening tool to evaluate the impacts of wastewaters on water quality and potential toxicity from 8 townships in Canada. In the context of municipal wastewaters management, the identification of toxic municipal effluents could help efforts to improve existing treatment strategies or contribute to the development of new ones. This rapid screening tool could also serve to reduce of bioassays using fish or other vertebrates for ethical considerations. Municipal wastewaters before (influent) and after 7 different treatment processes (effluents) were concentrated on a C18 reverse phase columns (ethanol elution) for the Per assay. An attempt was made to understand the contribution of population size and the performance of various treatments processes to alleviate changes in Per assay and the DNA protection effects.

2 Methods

2.1 Sample preparation

All reagents were purchased from Sigma Chemical Company (On, Canada) at the highest purity available. Horseradish peroxidase (er) and serum bovine albumin were prepared at 1 mg/mL in phosphate buffered saline (PBS: 140 mM NaCl, 5 mM KH2PO4 and 1 mM NaHCO3, pH 7.5) and kept for non-longer than one week at 4 °C in the dark. For DNA, 10 mg of salmon sperm DNA was dissolved in 10 mL of 0.25 X PBS, heated at 70 °C for 10 min to ensure complete dissolution and stored at 4 °C. The reagent dihydrofluorescein diacetate (DHFDA) was dissolved at 1 mg/mL in 10% ethanol and diluted to a working solution at 10 µg/mL in PBS buffer at the day of analysis. Hydrogen peroxide solutions were prepared at 1% concentration and stored at 4 °C for non-longer than one week.

The untreated (influents) and treated effluents from 8 townships of various treatment process and sizes in Canada were 24 h composites and collected for 3 days (Table 1). The wastewaters (1 L) were transported back to the laboratory and stored at 4 °C in the dark. The influent and corresponding effluent samples were filtered on 0.45 µm pore cellulose filter and 500 mL was passed through a reverse phase C18 cartridge (500 mg; Sulpeco). After washing with MilliQ water (10 ml), the material was eluted with 5 mL ethanol and concentrated to 1 mL under nitrogen stream (500 X concentrate). The ethanol samples were stored at -20 °C until analysis.

Table 1 City characteristics
Full size table

2.2 Wastewater extract characteristics

The dissolved organic matter (DOC) in the ethanol extract was determined by the spectrometric methodology [9]. The DOC levels were estimated based on the following relationship: DOC (mg/L) = 0.45A + 1 for organic-rich surface waters. The levels of plastic-like substances associated to the organic matter were determined by the copper fluorescence quenching methodology [10]. The decreased in fluorescence for humic/fulvic acids (265 nm excitation/463 nm emission), polypropylene-derived materials (250 nm excitation/324 nm emission) and polyvinyl chloride /polystyrene-derived materials (295 nm excitation/411 nm emission) were determined before and after the addition of 0.1 volume of 500 µM of CuCl2. The difference between fluorescence without Cu – fluorescence with Cu was calculated and standard solutions of polystyrene (20 nm diameter) and polypropylene (100 nm diameter) were used for calibration. The data were then normalized to the DOC in the samples.

The levels of light, medium and heavy polyaromatic hydrocarbons were determined in the ethanol extracts using fixed wavelength fluorometry [11]. Briefly, 100 µL of ethanol extracts were placed in dark 96-wells microplate and fluorescence measured at 290 nm excitation/340 nm emission (light PAhs: naphthalene), 325 nm excitation/370 nm emission (medium Pahs: pyrene) and 385 nm excitation /440 nm emission (heavy Pahs: benzo(a)pyrene) using a fluorescence microplate reader (Neo-2 Synergy, Biotek Instruments, USA). Standard solutions of the Pahs were used for calibration. Recovery of each Pahs size class were between 80–95% in the C18 extraction columns. The levels of polystyrene nanoplastics (PsNPs) were determined using a molecular rotor probe 9(dicyanovinyl)-julolidine as previously described [12]. Briefly, 10 µL of the effluent extract was mixed with 190 uL of 10 uM of the probe (diluted in MilliQ water) and fluorescence were taken at 450 nm excitation and 620 nm emission (Neo-2, Synergy-4, Biotech Instruments, USA). Standard solutions of polystyrene nanoparticles (50 nm Polyscience, USA) were used for calibration. Solvatochromatic analysis of the ethanol extracts to detect plastic-like substances using the Nile red methodology were also performed [13]. Briefly, 25 µL of the ethanol extracts were mixed with 225 µL of 50 µM Nile red (in PBS buffer) and the emission spectra were recorded between 520–700 nm at 485 nm excitation. The first derivative of the spectra revealed that PsNPs emitted at 600–620 nm and this signal was taken as a measured of plastic-like materials. The data were expressed as relative fluorescence units (600 nm) corrected for the organic matter contents. The levels of melamine were determined using a nano-gold plasmonic sensor [14]. Briefly, citrate coated nano-gold (10 nm diameter, NanoComposix, USA) were centrifuged at 20 000 × g for 10 min and the pellet resuspended in 0.001% Triton X-100. A 10 µL sample of the extract was mixed in 100 µL of nano-gold suspension, mixed for 5 min and the absorbance was taken between 500–700 nm (Neo-2, Synergy IV, Biotech Instruments, USA). The aggregation ratio was measured at 650 nm/520 nm. Blanks consisted of ethanol and standard solutions of melamine was used for calibration. The data were expressed as µg melamine/dissolved organic carbon content.

2.3 In vitro peroxidase assay

The in vitro peroxidase (Per) assay was based on a previous methodology for industrial effluents toxicity screening [6]. The Per assay principle is explained in Fig. 1, where Per and albumin are exposed to municipal effluent samples. DNA could be added to seek out interactions with municipal wastewaters extracts. The reagent DHFDA was used instead of luminol for fluorometric detection at 485 nm excitation and 530 nm emission for fluorescein at a final concentration of 1 µg/mL in PBS. The reaction media contained horseradish Per and albumin at 0.1 µg/mL in PBS in a total volume of 160 µL followed by the addition of 10 µL of the ethanolic extract and incubated for 5 min. After this period, 20 µL of 1 µM DCFDA and 0.01% H2O2 were added. The reaction was allowed to proceed at 25 °C for 30 min with readings taken at each 3 min for fluorescein evaluation (excitation 485 nm/emission 530 nm). The same procedure was repeated with the ethanol extract pre-incubated for 5 min with 1 µg/mL DNA to determine the influence of DNA on Per reaction rates. Blanks consisted of ethanol solvent (from the same amount of MilliQ water C-18 column and ethanol elution) and DNA in the presence of Per, albumin, H2O2 and DHFDA (Fig. 1). The DNA protection index is defined as: Per activity with DNA/Per activity. Blanks consisted of ethanol only and CdNO3 was used as a negative control (1 µg/L CdNO3 decreases Per activity by 30–40%). This Cd concentration is in the same range to trout toxicity (LC50 between 0.7–3 µg/L) for rainbow trout embryos and larvae [15]. Controls consists of MilliQ water in the same bottles for effluent collection and was extract on C-18 following the same procedure (500 × concentrate and ethanol elution). Blanks consisted of ethanol blanks and bottle controls to ensure no cross contamination from the wastewaters collection methods.

Fig. 1
figure 1

Principle of the in vitro peroxidase assay. Peroxidase catalyze the oxidation of organic (electron donor) compounds for the elimination of H2O2. A secondary reaction involves the peroxidase-oxidase reaction where O2 and NADH replaces H2O2 and electron donor compounds. The inhibition of peroxidase leads to the accumulation of H2O2, which rapidly produce toxicity and was associated to fish mortality [6]. Conversely, the activation of peroxidase leads to depletion of electron donors (e.g. ascorbic acids, glutathione/amino acids, DNA, RNA) and produce oxidative stress on the long-term. The DNA protection assay is calculated by the ratio (Per with DNA/Per alone) and indicates oxidative mediated affinity to DNA

Full size image

2.4 Data analysis

The in vitro exposure experiments to the various influents/effluents from 8 cities were repeated three times in the absence and presence of added DNA to determine DNA protection of the Per reaction. The population size of each (anonymous) cities and the various types of wastewater treatment were as follows (Table 1): aerated lagoon (Lag), lagoon with facultative aeration (LagF), advanced biofiltration (Adv), biological filtration (Bio), secondary activated sludge (Sec), secondary membrane bioreactor (SecM) and primary physico-chemical treatment (Prim). The untreated effluents were denoted by None. In this study the influence of treatment processes was obtained from cities differing in population sizes. To understand the influence of population size and the effluents properties from different treatment scenarios, the influents properties were first examined by population size (and were then examined by covariance analysis (ANCOVA) on log-transformed data with effluents from each treatments as the main variable and population size as the covariate. This analysis provides a statistical means to identify differences between treatment processes controlling for population size effects. Critical differences between the absence (Influent) of wastewater treatments and the following the 7 different treatment processes were determined by the LSD test. Significance was set at α = 0.05. The relationships between the endpoints were examined using hierarchical tree analysis using the Pearson moment correlation coefficient as the measure of distances (1−r). Discriminant function analysis was also used to seek the difference of wastewater properties before and after the treatment processes. All statistical analyses were performed using the Statsoft software package version 13.2 (USA).

3 Results

In this study, the comparison of various treatments is complicated by varaible population sizes across the examined cities. To mitigate these effects, the data for untreated wastewaters (influents or Inf) were examined with increasing population size to seek out population effects on incoming raw wastewater systems. We found are strong correlation between population size and flow rates (r = 0.94; p < 0.001) suggesting concordance between the population size and effluent flow rates. For comparisons between the treatment processes (and anonymous cities), an analysis of covariance was performed using population estimates as the covariate and the treatment types as the main variable. The basic chemical properties of the wastewaters was provided in Table 1S. The population densities ranged from 1200–1800000 inhabitants and cities 7 and 8 were the most populated, industrialized and emitters of TSS. The pH values ranged between 6.56 to 7.88 and total ammonia values ranged from 0.2 to 21.3 mg/L. The dissolved organic carbon contents (DOCs) were examined (Table 1). ANCOVA revealed that both treatment processes (p = 0.001) and population size (p < 0.001) were significantly related to DOCs. The analysis revealed that the LagF, Prim, Sec and SecM DOCs levels were significantly lower than the influents. In respect to population effects, the DOCs levels increased with population size (r = 0.42; p = 0.01). The relative levels of humic and fulvic acids (HA/FA) were also examined (Table 2). The analysis revealed that only the treatment types was significant with the SecM, Prim effluents showing lower levels compared to untreated influents. The levels of HA/FA were significantly increased in the Bio-treated effluents compared to the untreated influents. This suggests a greater input of natural organic matter from water sources. The levels of light and heavy Pahs were significantly affected by both population size and treatment processes while med Pahs were only affected by the treatment processes. The lights PaHs were significantly lower for the Sec, LagF and SecM effluents compared to untreated influents suggesting photo and biodegradation since the resident times are generally longer with these treatments (7–20 days in general). The heavy Pahs were significantly reduced by Prim, LagF, and SecM effluents compared to the Inf sites. In respect to population size, the levels of light and heavy Pahs increased with the population size at r = 0.3 and r = 0.54 respectively. For Med Pahs, the levels were significantly lower for lagF, SecM effluents but higher at the Bio effluent. The levels of melamine were also determined in these samples. The analysis revealed that only the treatment types (p < 0.01) influenced the levels of melamine (Table 2). The levels of melamine were significantly higher in the Lag, Sec and SecM effluents.

Table 2 Relative levels of light, medium and heavy PAHS and melamine in wastewaters
Full size table

In this study, the levels of plastic related compounds were examined in the dissolved organic matrix (Table 3). The levels of polystyrene nanoplastic particles (PsNPs) were significantly influenced by the population levels only (treatment types at p = 0.12; population at p < 0.001). The levels of PSNPs were significantly associated with population size (r = 0.64) and the treatment processes were seemingly invariable towards PSNPs. PSNPs were significantly related with TSS for population (r = 0.37). This suggests PsNPs are associated to the TSS fraction in wastewaters. For polypropylene-like materials, the analysis revealed that only the treatment types were significant (treatment at p = 0.001; population at p = 0.145). The levels of PP-like substances were significantly removed for the Bio and Sec effluents compared to the Inf. In respect to PS/PVC-like substances, only the treatment processes significantly differed (ANCOVA, treatments p < 0.001, population p = 0.34). In terms of untreated influents, the levels were somewhat higher in lower city population (r = 0.25; p = 0.05). The extracts were examined for the presence of plastic-like materials based on the solvatochromic properties of NR (Table 3). PsNPs emit at 600–620 nm in the presence of NR in the ethanol extracts and considered a signal for plastics materials. The analysis revealed that the treatment types (p < 0.001) only affected the levels of NR signal with no influence of population size (p = 0.32). NR signals were higher in the Adv, Lag, LagF, Sec and SecM treatments.

Table 3 Plastic nanoparticles and related materials
Full size table

The activity of Per was measured in the presence of various influents and effluents in the presence and absence of added DNA (Fig. 2) The analysis revealed that both the population size and treatment types significantly influenced Per activity (p < 0.001 for both variables) (Fig. 2A). The activity was significantly changed by all treatment types considering population size for Adv (decreased activity), Lag (increased), Prim (decreased), Sec(increased), SecM (increased) effluents following correction for population compared to controls. Indeed, activity in Per was positively correlated with population number (Fig. 2B, r = 0.50; p < 0.001). The DNA protection potential was significantly influenced by the treatment types (Fig. 3). Although most cities (by population number) produced DNA protection, suggesting a general release of genotoxic compounds, no trends were observed with population size. The following treatments increased the DNA protection potential relative to controls: Adv, Bio, Prim, Lag and the influents (no treatment).

Fig. 2
figure 2

Peroxidase activity in influents and effluents extracts. The influents and effluents were extracted on a C18-SPE columns and eluted with ethanol (corresponding to 500X). The extracts (50 X) were then tested with the peroxidase assay for the untreated effluents (A) and different wastewater treatment (WWT) type (B). The solid lines correspond to the control activity of peroxidase with the dotted lines corresponding to normal variation of 15%. In B, none is the mean value of the untreated influents in A. The star symbol * indicates significant difference from the control activity

Full size image
Fig. 3
figure 3

DNA protection assay of the peroxidase activity assay. The influents and effluents were extracted on a C18-SPE column and eluted with ethanol (corresponding to 500X). The extracts (50 X) were then tested with the peroxidase assay in the presence of DNA for the untreated effluents (A) and different wastewater treatment (WWT) type (B). The DNA protection was defined as Peroxidase with DNA (0.1 µg/mL) /peroxidase without DNA. The solid lines correspond to the absence of effects of DNA and the dotted lines corresponding to normal variation of 15%. In B, none is the mean value of the untreated influents in A. The star symbol * indicates significant difference from the control (absence of DNA effects)

Full size image

In the attempt to gain a more global view of the various endpoints, a hierarchical tree analysis and discriminant function analyses were performed (Fig. 4A and B). For the hierarchical tree analysis, the relative distance between the various endpoints was based on the correlation coefficient (1−r). The Per activity assay was closely related with the DNA-Per assay and were both related with the antioxidant melamine levels. The levels of PsNPs were closely related with heavy Pahs while light/medium Pahs were closely related with the dissolved organic content (DOC). The relative levels of HA/FA acids were closely related to PS/PVC-like substances. Population size was significantly related to Pahs, DOC and PSNPs. In the attempt to understand the performance of the treatment types, a discriminant function analysis was performed (Fig. 4B). The mean classification was 88% and the following markers had the highest factorial weights: melamine, Per, light/med Pahs and PS-like substances. The Adv, Bio, SecM and Sec produced the most difference between the influents and effluents based on the above endpoints. The Lag and Prim produced less difference between the untreated influents and treated effluents suggesting that the effluents were closely related in properties to the untreated effluents.

Fig. 4
figure 4

Hierarchical tree and discriminant function analyses of influents and effluents contaminants. Tree analysis was provided using the Pearson moment correlation coefficient (1−r) (A). The dotted line represents the significance limit of the correlation coefficients. Discriminant function analysis is provided to seek out changes by the various treatment processes, before and after treatment (B). The untreated influents and effluents are highlighted as orange and green respectively

Full size image

4 Discussion

The in vitro Per assay was first introduced as a rapid and inexpensive screening tool for water quality assessments for urban, agriculture and industrial wastewaters in aquatic ecosystems [7, 16]. Application of this assay directly on wastewaters could also pick up the presence of microorganisms if the samples are not filtered or extracted first [17]. When this assay was used on filtered (0.2 µm) industrial effluents, Per inhibitions were significantly associated trout survival and Ceriodaphnia dubia survival and reproduction data [6] giving a toxicological meaning for this in vitro assay and its potential use as a NAMs for toxicity testing. In some cases, the activity of Per could be observed. In the case of decreased activity, H2O2 could increase at toxic levels very quickly. In the case of increased activity, H2O2 levels are reduced at the expense of the antioxidants acceptor molecular depleting antioxidant levels or forming potentially oxidized by products. The composition of this assay is strikingly simple with Per enzyme and albumin in solution with H2O2 and DHFDA as the antioxidant molecule. In addition, DNA could be added to identify samples that prevent changes in Per activity, hence DNA protection assay. On the one hand, inhibition of Per activity was associate to acute toxicity from H2O2 build-up in cells. On the other hand, the increase in Per activity, while maintaining H2O2 at safe levels, involves the accumulation of harmful oxidized products on the long-term basis. The presence of hydroperoxides oxidized polyunsaturated lipids, are also likely [18]. These peroxidases usually use glutathione as co-substrates and lipid hydroperoxide instead of H2O2. Municipal effluents, and sludge flocs were shown to contain unsaturated lipids, which could be oxidized especially in those station plants using aeration with sunlight or UV treatments [19]. Moreover, biofilms rich in extracellular polymeric substances were also particularly rich in unsaturated fatty acids. This is in keeping with lagoons using secondary aeration or UV treatment (Sec, SecM and AL), which showed the highest levels of induction in Per activity.

Oxidative stress often involves the activation of catalase, superoxide dismutase and Per leading to the accumulation of oxidative damage (lipids, proteins and DNA) on the long-term [20, 21]. For example, increased Per activity was often associated with oxidative damage such as lipid peroxidation and DNA damage [22, 23]. This suggests that sustained Per activity although contributing to lower H2O2 levels could decreases the antioxidant capacity of cells on the long-term (Fig. 2). The production of H2O2 by superoxide dismutase (SOD) is usually controlled by catalase and peroxidases. While catalase eliminates H2O2 into O2 and H2O, Per involves the co-oxidation of endogenous and exogenous compounds. While xenobiotics could be oxidized by Per, the enzyme co-oxidizes various endogenous ligands such as vitamin C, glutathione and perhaps endogenous nucleotides (DNA oxidation), polyunsaturated lipids (lipid peroxidation) and proteins (carbonylation). It is not clear whether H2O2 alone or in combination with Per contributes mainly to the formation of DNA damage and lipids. It is assumed, that H2O2 alone directly oxidizes lipids and DNA in cells. For example, aromatic hydrocarbons could form DNA adducts leading to strand breaks in the presence of H2O2 and Per [4]. In our controls, Per activity was not significantly changed (albeit somewhat lower) in the presence of DNA suggesting that DNA was not a direct substrate for Per (Fig. 3). More research would be needed to clarify this point. It is noteworthy that melamine levels were significantly correlated with Per activity with or without added DNA. Melamine has 2 oxidation sites leading to ammelide and cyanuric acid [24], where melamine-cyanuric acid complex can lead to DNA damage and renal damage [25]. Interestingly, single exposures of either cyanuric acid or melamine in human embryonic kidney 293 cells did not lead to DNA strand breaks but co-exposure to them produced DNA strand breaks. Melamine could also act as an enhancer in the Per reaction where melamine would act as secondary substrate the peroxidation of DHFDA [26]. Enhancers of the Per oxidation usually are nitrogen aromatic compounds such as aniline, phenothiazine and benzidine. If this holds true, melamine (and its oxidized products ammelide and cyanuric acid) could accept one electron from Per and the melamine radical would, in turn, oxidize the substrate DHFDA. It was shown that melamine could increased the peroxidase properties of gold nanoparticles in the presence of H2O2 and tetramethylbenzidine [26], which supports this hypothesis.

In a previous study with industrial effluents, DNA protection of the Per reaction (increased DNA-Per activity compared to Per activity alone) was associated in 70% of cases with increased DNA (SOS) repair activity in bacteria [6] making it a cheap and rapid screening test for genotoxicity. Although the maximum reduction on Per activity (0.2 fold of the control activity) was observed at 50 X concentrations, toxic effects are unlikely at effluent concentrations < 100% (1 X). The Per assay was initially developed with industrial wastewaters but the lethal toxicity to fish was not included for municipal wastewaters. Trout toxicity for these municipal effluents during regulatory compliance testing were done (https://open.canada.ca/data/en/dataset/471736af-e236-444a-9888-b4d99052c927) and corroborates our finding with Per assay towards fish toxicity i.e., no lethal toxicity found in these samples at concentrations < 1X. This provides evidence that the Per assay did not produce any false negative municipal effluents in respect to rainbow trout toxicity. Although the Per assay was significantly related with trout and Ceriodaphnia dubia survival tests (multiple regression r = 0.65; p = 0.02), positive results with Per and or the DNA protection assay should be confirmed at the fish/daphnid levels to further confirm the observed trend observed for industrial effluents [6]. Nevertheless, increased Per activity by enhancers could present a more long-term stress since these compounds are usually (geno)toxic in addition to the presence of endocrine disruption in domestic wastewaters. It is noteworthy that the untreated influents inhibited Per more strongly in smaller populated townships suggesting perhaps some limitation of raw wastewaters collection infrastructure (Fig. 2). In respect to sublethal effects, SOD gene expression, involved in the production of H2O2, was examined in rainbow trout hepatocytes exposed to municipal effluent extracts from 12 Canadia cities [27]. Over 75% of the effluents produced changes in SOD gene expression indicating the production of reactive oxygen species and supporting the knowledge that municipal effluents produce oxidative stress in organisms. Compared to the untreated influents, SOD gene expression was increased in 3 of the tested effluents and that population and DOC were significantly related to SOD gene expression.

It was noteworthy that the DNA protection assay revealed significant protection in 5/8 (63%) of the wastewaters providing evidence that municipal effluents are often genotoxic to aquatic organisms (Fig. 3). This observation corroborates previous findings showing that municipal effluents from a primary and advanced biofiltration with secondary UV-treatment were equally genotoxic in fathead minnows exposed for 3 months to these effluents [28]. This suggests that genotoxic compounds were impervious to differing water treatments and that the occurrence of DNA damaging compounds is relatively common in treated wastewaters. Genotoxicity calculations of the effluent of the Montreal (Quebec, Canada) revealed that it releases over 31 kg of benzo(a)pyrene equivalents per day using a bacterial DNA-repair assay- SOS Chromotest [29]. Moreover, the data also revealed that 90% of the genotoxic loadings was nonindustrial in origin i.e., from domestic wastes and street runoffs. In a following study with mussels caged to the municipal dispersion plume of the City of Montreal for 1 month, DNA strand breaks were detected in hemocytes using the Comet assay [30]. The genotoxicity of treated and untreated municipal effluents was also observed in Elliptio complanata mussels hemocytes exposed in vitro to municipal effluent extracts [31]. The study revealed that genotoxicity persists following treatment in other municipalities across Canada suggesting the continuous release of genotoxic compounds in the aquatic environment. The release of toxic chemicals in the aquatic ecosystem are expected to increase in these times of global warming as the frequency and severity of rainfall events increases. The development of quick and inexpensive tests to assist the regulatory and waste management communities to preserve water quality are therefore needed. In the context of reducing or replacing fish tests for ethical values for routine monitoring of wasters and rain runoffs, these quick and rapid tests could be of value as alternative methods. However, more comparative information between the Per assay and trout survival is required for real environmental mixtures.

5 Conclusions

The water quality of municipal wastewaters was examined with the Per activity and DNA protection assays to highlight potential toxic (Per inhibitors or activator) and genotoxic chemicals during the rapid screening of wastewaters or leachates. The study revealed that Per activity of wastewaters influents were influenced by population size and the wastewater treatment types. The more advanced treatments (biofiltration, membrane biofiltration, secondary aeration) produced stronger changes compared to the corresponding untreated influents. Per activity changes were also associated to the DOC, Pahs and plastic-like substances and nanopplastics levels in the effluents. A positive relationship was observed with melamine a Per activity suggesting that other reduced substances could act as substrates for Per as well and could explain, in part, interactions with DNA (genotoxicity). In conclusion, the Per assay represents a cost effective tool for the rapid screening of various complex mixtures such as industrial and municipal wastewaters. The data show that municipal effluents are generally not acutely lethal based on inhibitions of the Per activity but could in some instances increase Per activity leading to more long-term toxicity by the accumulation of oxidizes endogenous substrates. The DNA protection assay revealed that 60% of the effluents have a genotoxic potential providing some evidence on the occurrence of genotoxic compounds in aquatic ecosystems by municipal wastewaters.