Effluents from sewage treatment plants contain a mixture of micropollutants with the potential of harming aquatic organisms. Thus, addition of advanced treatment techniques to complement existing conventional methods has been proposed. Some of the advanced techniques could, however, potentially produce additional compounds affecting exposed organisms by unknown modes of action. In the present study the aim was to improve our understanding of how exposure to different sewage effluents affects fish. This was achieved by explorative microarray and quantitative PCR analyses of hepatic gene expression, as well as relative organ sizes of rainbow trout exposed to different sewage effluents (conventionally treated, granular activated carbon, ozonation (5 or 15 mg/L), 5 mg/L ozone plus a moving bed biofilm reactor, or UV-light treatment in combination with hydrogen peroxide). Exposure to the conventionally treated effluent caused a significant increase in liver and heart somatic indexes, an effect removed by all other treatments. Genes connected to xenobiotic metabolism, including cytochrome p450 1A, were differentially expressed in the fish exposed to the conventionally treated effluents, though only effluent treatment with granular activated carbon or ozone at 15 mg/L completely removed this response. The mRNA expression of heat shock protein 70 kDa was induced in all three groups exposed to ozone-treated effluents, suggesting some form of added stress in these fish. The induction of estrogen-responsive genes in the fish exposed to the conventionally treated effluent was effectively reduced by all investigated advanced treatment technologies, although the moving bed biofilm reactor was least efficient. Taken together, granular activated carbon showed the highest potential of reducing responses in fish induced by exposure to sewage effluents.
Overall design: The fish exposure was carried out in March 3-17, 2008 at Henriksdal STP (Stockholm, Sweden) using a large scale pilot plant with parallel treatment lines under ethic permits approved by the local animal committee in Gothenburg (permission no. 36-2007). A thorough description of the different treatment steps (both conventional and advanced) and the experimental setup during the exposure (Fig. 1) has been previously described elsewhere (Samuelsson et al., 2011). In brief, the raw influent underwent conventional activated sludge treatment to generate a reference effluent (Conventional Activated Sludge; CAS). This includes mechanical treatment (removal of large objects using grids with 3 mm column width); primary sedimentation; biological treatment (activated sludge aged 20 days); secondary sedimentation; sand filter. The additional advanced treatment processes were granular activated carbon (GAC), ozonation at 5 mg/L (OZ5) or 15 mg/L (OZ15), ozonation 5 mg/L followed by a moving bed biofilm reactor (MBBR) and irradiation by ultraviolet radiation (0.75 Wh/L) plus hydrogen peroxide (10 mg/L; UH). Two aquaria were used for positive controls. As the tap water available is not as well accepted by the fish, and to better reflect the organic content and overall chemistry of surface water, the control aquaria (TW1 and TW2, together TW) contained carbon-filtered tap water supplemented by 2% of conventionally treated reference effluent. We consider evaluation of implementation and running costs to be outside the scope of this paper.
Juvenile rainbow trout (Oncorhynchus mykiss; 113±23 g; mean bodyweight ± standard deviation) of both sexes were obtained from Antens Laxodling AB (Alingsås, Sweden). Rainbow trout is a commonly used species for this type of studies and our lab have ample experience using this species. Furthermore, rainbow trout of sufficient size for metabolomic studies (Samuelsson et al., 2011) can easily be obtained. Finally, the use of trout provides the opportunity for comparisons with other studies (e.g. Gunnarsson et al., 2009 and Albertsson et al., 2010). During an acclimatization period of four days on site, the fish were kept in tanks containing carbon-filtered tap water + 2% reference effluent. On the onset of the experiment, the fish were randomly distributed among eight 100 L aquaria (n=25/aquarium), supplied with treated effluent from the different parallel treatment lines. Aquaria duplicates were used for the controls. During both the acclimatization period as well as the 14 day exposure, the aquaria were aerated and the fish were not fed to reduce variability due to differences in food intake. It should be stressed that trout easily cope with food deprivations for several weeks (Kullgren et al., 2010). Measurements of temperature (13.5±0.5°C), oxygen saturation (97±3%), pH (7.4±0.3), conductivity and flow rate were monitored 5 d per week throughout the exposure. In addition, to evaluate differences in removal efficiency between the treatments, total organic carbon (TOC), 7-day biochemical oxygen demand (BOD7), suspended solids (SS), total nitrogen (Tot-N) and total phosphorus (Tot-P) were measured in all effluents (Samuelsson et al., 2011).
The sampling took place during two consecutive days (day 14 and 15) due to the large number of fish. For each day, an equal number of fish from each treatment was sampled in random order and killed by a blow to the head. The fish were weighed, their fork length was measured and their sex was determined by macroscopical observation of their gonads.
Liver samples were collected for several different studies (hence sample size limitations), frozen on dry ice and stored at -70°C until analysis. Homogenization of the frozen liver tissue was carried out using Tissuelyser (Qiagen, Hilden, Germany) and hepatic total RNA was extracted and purified using QIAcube and RNeasy® Plus Mini Kit (Qiagen). The RNA quantity and quality were assessed by spectrophotometric measurements with the Nanodrop 1000 (NanoDrop Technologies, Wilmington, DE). To ensure that no degradation had occurred, the isolated RNA was analyzed using Experion automated electrophoresis (Bio-Rad, Hercules, CA).
A 15k rainbow trout gene expression microarray was designed for the RT analyzer platform (febit, Heidelberg, Germany) by using The Institute for Genomic Research (TIGR) Rainbow Trout Gene Index (RTGI) database version 7.0 (http://compbio.dfci.harvard.edu/tgi/). Details on the probe design strategy, but for eelpout (Zoarces viviparus), and transcript selection strategy are described elsewhere (Kristiansson et al., 2009; Cuklev et al., 2011). However, in the present study, singletons and non-annotated expressed sequence tags (ESTs) were replaced by newly well-annotated ESTs in rainbow trout. When available, transcripts at GenBank (/nucleotide) were used. In our lab, results from similar microarrays using the same platform have shown good correlation with quantitative polymerase chain reaction (qPCR) data (Gunnarsson et al., 2009b; Kristiansson et al., 2009; Cuklev et al., 2011; Lennquist et al., 2011).
To reduce variation in estrogen-responsive genes, only males were used for the subsequent gene expression analyses. Biotinylated antisense RNA (aRNA) was synthesized using MessageAmp™ II-Biotin Enhanced Single Round aRNA Amplification Kit (Ambion®). The aRNA samples (20 µg) were vacuum dried in a vacuum centrifuge, dissolved in 10 µl water and fragmented according to the manufacturer’s protocol. The following steps described in this subchapter were all performed by febit. Oligonucleotide arrays were synthesized by photo-controlled in situ synthesis using the Geniom One system (febit). Each biochip consists of eight individually accessible micro channels, each of which is referred to as a microarray in this manuscript. Eight individual samples from each aquarium were included in the analysis. In total, 64 microarrays were analyzed. A customized protocol, described in detail elsewhere (Cuklev et al., 2011), was used for prehybridization and hybridization. All samples were randomly placed on the biochips, with one sample from each exposure on each biochip. Signals were detected using the internal CCD-camera system of the RT analyzer instrument (febit) and quantified using the Geniom Wizard software. Integration times were between 156 and 570 ms, determined automatically by the instrument software.
All microarray data processing and statistical calculations were performed in R-2.12.2 (www.r-project.org; R Development Core Team, 2010). The quality of pre- and post normalized arrays was verified with box- and MA plots. The data analysis was performed in the R-package LIMMA (Smyth, 2005). Data were normalized using the ‘quantile’ method. Moderated t-statistics and adjusted p-values of differential expression were calculated using the empirical Bayes model.
Less...