En vurdering av primære leverceller fra regnbueørret (Oncorhynchus mykiss ) sin egnethet som analyseverktøy for østrogenreseptor-agonister

Regulatory legislations on chemicals have in recent years become more stringent, requiring more toxicological evaluation. The increased toxicological testing have resulted in a stronger effort to implement the 3Rs (Refinement, Reduction and Replacement) through identifying suitable alternatives (non-animal) to animal testing. Evaluation of alternative methods such as cell-based in vitro methods (e.g. continuous cell lines, tissues slices and primary cultures) have shown to be promising, but there are currently few validated alternative bioassays for fish. The present work aimed to evaluate the primary rainbow trout (Oncorhynchus mykiss) hepatocyte model's suitability as an ecotoxicological tool in screening of single and mixtures of chemicals with modulatory properties on the estrogen receptor (ER)-activity and auxiliary endpoints such as cytotoxicity and biotransformation (e.g. aryl hydrocarbon receptor (AhR)-activity). The primary hepatocytes sensitivity, reproducibility and responsiveness were assessed for seasonal, donor and assay-related variability to address potential factors affecting the bioassay reproducibility. The hepatocyte model's suitability to characterise environmentally relevant concentrations of estrogens, anti-estrogens and organic compounds was assessed using both single endpoint and broad content approaches (global transcriptomics) after exposure to single compounds and mixtures of these with similar or dissimilar mode of action (MoA). Compounds estrogenicity and anti-estrogenicity were assessed in the fish hepatocytes using classical estrogen sensitive biomarkers (e.g. ERα and ER-mediated egg-yolk precursor vitellogenin (Vtg) and egg shell zona radiata (zrp)) complemented by determination of cytotoxicity (cell membrane stability and metabolic activity), AhR-mediated responses (ahr, cytochrome P450 1a (cyp1a), enzymatic activity of ethoxyresorufin-O-deethylase (EROD)). Characterisation of chemicals additional MoA in the cells were performed using novel analytical tools such as high-density oligonucleotide salmonid microarray in combination with quantitative real-time polymerase chain reaction (qPCR). Results from the studies demonstrated primary hepatocytes ability to facilitate detection of ER-mediated responses, biotransformation and cell growth-related gene expression by ER-agonists and acute toxic chemicals during short-term exposures (<96h). The bioassays ability to remain unaffected by variable parameters (robustness) and sensitivity was not affected by seasonal variations in ER sensitivity (Vtg gene and protein expression), but was dependent on individual donor-physiology variability in exposure studies with the model ER-agonist 17α-ethinylestradiol (EE2). The cell-batch variability was however reduced when optimal exposure time and data normalization was applied, yielding a concentration-dependent Vtg gene and protein expression. In addition to individual donor-physiology, bioassay related factors such as hepatocyte culturing conditions and normalization procedures were identified as possible cause to the observed variability. The variability within the bioassay may be reduced through protocol harmonization (e.g. same-species culturing conditions, cell density, media supplements), resulting in a more robust and reproducible assay for assessing ER active compounds. The primary hepatocytes suitability as a screening tool for ER-agonists were further displayed in measured classical biomarker genes, identifying ER signalling and its associated pathways as the main target of the ER-agonist in the cells. Exposure to the ER-agonist caused similar transcriptional responses in vitro as previously reported in vivo, suggesting the hepatocytes to facilitate relevant biomarker responses in screening of ER-active compounds. The suitability of the in vitro model in transcriptional and sub-cellular characterization of anti-estrogenic binary and ternary mixtures (AhR-agonist (β-naphtoflavone (BNF), ER-agonist (17 β-estradiol (E2)), partial ER-antagonist (hydroxytamoxifen (OHT))) were performed to better understand how compounds with dissimilar mode of action (MoA) contributed to combined anti-estrogenic effects. The findings displayed significantly increased anti-estrogenic effect in the combined mixture of BNF, OHT and E2, indicative of the individual compounds MoA to contribute to the total anti-estrogenic effect in the primary hepatocytes. The results suggested that the compounds in the mixture induce nuclear receptor-mediated cross-talk involving AhR-mediated transcription of increased cyp1a metabolism of the ER-agonist and binding and inhibition of further ER activity by the ER-antagonist. The combined mixture's anti-estrogenic effect is therefore suggested due to the compounds differently acting MoA as they have the same effect (e.g. reduction of Vtg protein) but through differently acting pathways. To further assess the primary hepatocyte bioassay ability to identify complex and previous un-described ER-agonists, a broad range of uncharacterised mixtures (UCM)-related compounds of naphthenic acids and hydrocarbons were screened for potential estrogenicity. Few compounds induced weak estrogenic activity in the primary hepatocytes as the majority of the tested compounds had a narcotic MoA and reached their water solubility before eliciting any ER-activity. Auxiliary endpoints such as EROD activity could not explain the compounds weak estrogenicity and further supported their narcotic MoA. The potential estrogenicity might however be masked by the compounds highly variable physico-chemical properties that may have affected their bioavailability in the in vitro system. The present work has demonstrated that the primary rainbow trout hepatocyte model is a versatile, multi-endpoint tool for screening ER-agonists using both single biomarker and global gene expression approach. The bioassay provided reproducible results that demonstrate its sensitivity, robustness and responsiveness in ecotoxicological screening of chemicals that modulate the activity of the ER and downstream cellular events. These findings may contribute to a better mechanistic understanding of well-characterised and novel MOA related to single and combined chemical exposures in the in vitro fish model. The global gene expression was a good unbiased tool when characterizing the MoA of ER-active chemicals in the primary hepatocytes as it unravelled relevant in vivo ER-mediated responses, hence displaying the model's potential to become a (eco)toxicological tool in ER-agonist screening.Regulatorisk kjemikalielovgivning har i de senere årene blitt strengere, noe som krever mer toksikologisk vurdering. Den økte toksikologisk testingen har ført til en sterkere innsats for å implementere de 3Rer (forbedring, reduksjon og erstatning) gjennom å identifisere egnede alternativer til dyreforsøk. Vurdering av alternative metoder som cellebaserte in vitro metoder (f.eks kontinuerlige cellelinjer, vev-skiver og primære kulturer) har vist seg å være lovende, men for tiden er det få validerte alternative metoder for fisk. Dette arbeidet evaluerer primære regnbueørret (Oncorhynchus mykiss) hepatocytters egnethet som et økotoksikologisk verktøy i screening av enkeltstoffer og blandinger av kjemikalier med modulerende egenskaper på østrogenreseptor (ER)-aktivitet med hjelp av endepunkter som cytotoksisitet, østrogenrespons og biotransformasjon (f.eks aryl hydrokarbon reseptor (AhR)-aktivitet). Sensitivitet, reproduserbarhet og reaksjonsevne for metoden ble vurdert ved å se på responsvariasjon knyttet til sesong, donorfisk og analysemessige variasjoner, og potensielle faktorer som påvirker bioassayets reproduserbarhet ble identifisert. Hepatocytt-modellens egnethet for å karakterisere miljørelevante konsentrasjoner av østrogener, anti-østrogener og organiske forbindelser ble vurdert ved bruk av både enkelt endepunkter og analyse genuttrykk etter eksponering for enkeltstoffer og blandinger av disse med tilsvarende eller ulik virkningsmekanisme (MoA). Østrogenisitet og anti-østrogenisitet av kjemikalieblandinger ble vurdert i fiskehepatocytter ved hjelp av klassiske østrogensensitive biomarkører (f.eks ERα og ER-mediert eggeplomme forløper vitellogenin (Vtg) og eggeskallkomponenten zona radiata (ZRP)) supplert med bestemmelse av cytotoksisitet (cellemembranstabilitet og metabolsk aktivitet), arylhydrocarbon reseptor (AhR)-medierte responser (Ahr, cytokrom P450 1a (CYP1A), og enzymatiske aktivitet til etoksyresorufin-O-deetylase (EROD)). Karakterisering av kjemikalier for ytterligere MoA i cellene ble utført ved bruk av nye analyseverktøy som oligonukleotid mikromatrise for laksefisk i kombinasjon med kvantitativ real-tid polymerase kjedereaksjon (qPCR). Resultater fra studien viste primære hepatocytters evne til å å påvise ER-medierte reaksjoner, biotransformasjon og cellevekst knyttet til genekspresjon etter korttidseksponering (