The lack of freshwater resources worldwide is partly caused by wastewater discharge that has not been properly or adequately treated. A further problem is emerging contaminants (CECs), which include mycotoxins, surfactants, flame retardants, industrial additives, medicines, personal care products (PCPs). The main consequences of CECs on the environment are acute hazardous events, however factors like bioaccumulation and persistence may also have long-term unfavourable effects. Pharmaceuticals are everywhere and are constantly released into the environment. Their unique traits (such as high molecular weight, lipophilicity, persistence, and bioaccumulation) increase their toxicity and harmful consequences on ecosystems. There are various methods of assessing the toxicological effects of these pollutants on receptors. In silico methods, which do not involve laboratory testing but nevertheless rely on experimental data, are a quite recent and innovative study field. In silico approaches relate molecular structural characteristics to their function and any possible impacts. In fact, it is assumed that molecules with similar chemical structures share biological activities and that the link between structure and activity depends on the chemical constitution. In this thesis project, I evaluated the useful parameters for the acute fish toxicity assessment of pharmaceutical compounds in fish using one of the in silico approaches called read-across. Read-across is used for substances whose toxicological profile is unknown and the behaviour of the target molecules is evaluated using similar molecules. To analyse the useful parameters for the toxicity assessment using this method, I first prepared the necessary dataset and searched for structural alerts. After assessing the reliability of the alerts and investigating their mechanism of action, the fragments were implemented in VERA, a new read across software. I investigated other parameters useful for the read-across approach, i.e. molecular descriptors describing the physicochemical properties of molecules. This allows us to explore other important aspects beyond structural similarity. As a result, I have found structural alerts that show high reliability rates, which are useful in characterising the toxicological profile of molecules. By reasoning about molecular descriptors, it was also seen that molecules that are structurally similar to the target and share structural alerts are sometimes not representative of the target, as they differ significantly in certain physicochemical properties, with strong implications on toxicity assessment.
La mancanza di risorse di acqua dolce in tutto il mondo è in parte causata dallo scarico di acque reflue che non sono state trattate in modo corretto o adeguato. Un ulteriore problema è rappresentato dai contaminanti emergenti (CEC), che comprendono micotossine, tensioattivi, ritardanti di fiamma, additivi industriali, farmaci e prodotti per la cura personale (PCP). Le principali conseguenze dei CEC sull'ambiente sono eventi pericolosi acuti, ma anche fattori come il bioaccumulo e la persistenza possono avere effetti negativi a lungo termine. I farmaci sono ovunque e vengono costantemente rilasciati nell'ambiente. Le loro caratteristiche uniche (come l'elevato peso molecolare, la lipofilia, la persistenza e il bioaccumulo) ne aumentano la tossicità e le conseguenze dannose sugli ecosistemi. Esistono diversi metodi per valutare gli effetti tossicologici di questi inquinanti sui recettori. I metodi in silico, che non prevedono test di laboratorio ma si basano comunque su dati sperimentali, sono un campo di studio piuttosto recente e innovativo. Gli approcci in silico mettono in relazione le caratteristiche strutturali delle molecole con la loro funzione e gli eventuali impatti. Si presume, infatti, che molecole con strutture chimiche simili condividano attività biologiche e che il legame tra struttura e attività dipenda dalla costituzione chimica. In questo progetto di tesi, ho valutato i parametri utili per la valutazione della tossicità acuta dei composti farmaceutici nei pesci utilizzando uno degli approcci in silico chiamato read-across. Il read-across viene utilizzato per sostanze di cui non si conosce il profilo tossicologico e il comportamento delle molecole target viene valutato utilizzando molecole simili. Per analizzare i parametri utili per la valutazione della tossicità con questo metodo, ho innanzitutto preparato il set di dati necessario e ho cercato gli allarmi strutturali. Dopo aver valutato l'affidabilità degli allarmi e indagato sul loro meccanismo d'azione, i frammenti sono stati implementati in VERA, un nuovo software di read across. Ho studiato altri parametri utili per l'approccio read-across, cioè i descrittori molecolari che descrivono le proprietà fisico-chimiche delle molecole. Questo ci permette di esplorare altri aspetti importanti oltre alla somiglianza strutturale. Di conseguenza, ho trovato segnalazioni strutturali che mostrano alti tassi di affidabilità, utili per caratterizzare il profilo tossicologico delle molecole. Ragionando sui descrittori molecolari, si è anche visto che le molecole strutturalmente simili al bersaglio e che condividono gli allarmi strutturali a volte non sono rappresentative del bersaglio, in quanto differiscono significativamente in alcune proprietà fisico-chimiche, con forti implicazioni sulla valutazione della tossicità.
Valutazione della tossicità acuta nel pesce con metodi di read-across
DI CUGNO, VIVIANA
2021/2022
Abstract
The lack of freshwater resources worldwide is partly caused by wastewater discharge that has not been properly or adequately treated. A further problem is emerging contaminants (CECs), which include mycotoxins, surfactants, flame retardants, industrial additives, medicines, personal care products (PCPs). The main consequences of CECs on the environment are acute hazardous events, however factors like bioaccumulation and persistence may also have long-term unfavourable effects. Pharmaceuticals are everywhere and are constantly released into the environment. Their unique traits (such as high molecular weight, lipophilicity, persistence, and bioaccumulation) increase their toxicity and harmful consequences on ecosystems. There are various methods of assessing the toxicological effects of these pollutants on receptors. In silico methods, which do not involve laboratory testing but nevertheless rely on experimental data, are a quite recent and innovative study field. In silico approaches relate molecular structural characteristics to their function and any possible impacts. In fact, it is assumed that molecules with similar chemical structures share biological activities and that the link between structure and activity depends on the chemical constitution. In this thesis project, I evaluated the useful parameters for the acute fish toxicity assessment of pharmaceutical compounds in fish using one of the in silico approaches called read-across. Read-across is used for substances whose toxicological profile is unknown and the behaviour of the target molecules is evaluated using similar molecules. To analyse the useful parameters for the toxicity assessment using this method, I first prepared the necessary dataset and searched for structural alerts. After assessing the reliability of the alerts and investigating their mechanism of action, the fragments were implemented in VERA, a new read across software. I investigated other parameters useful for the read-across approach, i.e. molecular descriptors describing the physicochemical properties of molecules. This allows us to explore other important aspects beyond structural similarity. As a result, I have found structural alerts that show high reliability rates, which are useful in characterising the toxicological profile of molecules. By reasoning about molecular descriptors, it was also seen that molecules that are structurally similar to the target and share structural alerts are sometimes not representative of the target, as they differ significantly in certain physicochemical properties, with strong implications on toxicity assessment.È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per maggiori informazioni e per verifiche sull'eventuale disponibilità del file scrivere a: unitesi@unipv.it.
https://hdl.handle.net/20.500.14239/15143