STUDY OF LIPIDOMIC FINGERPRINT OF 3D HUMAN LIVER MICROTISSUES AS PREDICTION OF DRUG INDUCED LIVER INJURY

With more than 180,000 different molecular species, lipids form a vast and extraordinary class of biomolecules. Functioning as structural components of membranes, as medium for energy storage, as anchor for proteins, or as intra- and inter-cellular signaling molecules, lipids play different roles and are involved in several biological processes. The total lipid composition of an organism or specimen thereof is referred to as the lipidome and lipidomics, a subspecialty of metabolomics, is the study of lipidomes using the principles and techniques of analytical chemistry. Emerged in 2003, lipidomics represents an emerging field to investigate the lipidome of diseases’ epidemiology, with the aim of unravelling diagnostic biomarkers and new drug targets and of rationalizing toxicity effects. Indeed, alteration of lipid regulation can assist in the pathophysiology of diseases including diabetes, obesity, heart diseases, infectious diseases, or neurodegenerative diseases. Furthermore, different drugs induce lipid disorders, constituting major risk factors for hepatic injury. Drug-induced lipid disorders are often unpredictable in preclinical stages, and related adverse events have led to post-market withdrawal of drugs. Accordingly, there is a need for more predictive tools to assess hepato-toxicity risk in drug discovery. In according with the previous observations, we study the variability of lipid pattern in liver in presence of different drugs, in vitro condition. For this approach we used 3D InSightTM human liver microtissues, which are characterized by high sensitivity and stability and allows the hepatotoxicity studies in a “milieu” closely resembling the in vivo conditions. The microtissues were incubated with well-known drugs up to 14 days and lipids were extracted at different time (0,2,4,7,9,11,14 days). The lipid extraction from each sample was analysed by LC-MS/MS and lipid identification and quantification were performed by means of Lipostar software. From the more of 900 lipids identified, a lipid fingerprint composed of 300 selected lipids was generate. PCA analysis of the samples showed that lipid fingerprints changed with drug-mediate toxic effects and with incubation time. At the same time the metabolic profile of the investigate drugs was obtained. The results clearly indicate that lipidomic data can be used to estimate the concentration-toxic effect relationship. Moreover, this effect can be linked either with parent or metabolite compounds. This approach can be useful in the early stages of drug-discovery for characterizing the toxicological and other biological effects as well as the ADME properties of novel compounds.