A Metabolomics Investigation of Non-genotoxic Carcinogenicity in the Rat

Abstract

Non-genotoxic carcinogens (NGCs) promote tumour growth by altering gene expression which ultimately leads to cancer without directly causing a change in DNA sequence. As a result NGCs are not detected in mutagenesis assays. Whilst there are proposed biomarkers of carcinogenic potential, the definitive identification of non-genotoxic carcinogens still rests with the rat and mouse long term bioassay. Such assays are expensive, time consuming, require a large number of animals and their relevance to human health risk assessments is debatable. Metabolomics and lipidomics in combination with pathology and clinical chemistry were used to profile perturbations produced by 10 compounds which represented a range of rat non-genotoxic hepatocarcinogens (NGC), non-genotoxic non-hepatocarcinogens (non-NGC) and a genotoxic hepatocarcinogen. Each compound was administered at its maximum tolerated dose level for 7, 28 and 91 days to male Fisher 344 rats. Changes in liver metabolite concentration differentiated the treated groups across different time points. The most significant differences were driven by pharmacological mode of action, specifically by the peroxisome proliferator activated receptor alpha (PPAR-α) agonists. Despite these dominant effects, good predictions could be made when differentiating NGCs from non-NGCs. Predictive ability measured by leave one out cross validation was 87% and 77% after 28 days of dosing for NGCs and non-NGCs, respectively. Amongst the discriminatory metabolites we identified free fatty acids, phospholipids, triacylglycerols, as well as precursors of eicosanoid and the products of reactive oxygen species linked to processes of inflammation, proliferation and oxidative stress. Thus, metabolic profiling is able to identify changes due to the pharmacological mode of action of xenobiotics and contribute to early screening for non-genotoxic potential.

PPAR-pan activation induces hepatic oxidative stress and lipidomic remodelling

“…The PLS-DA model variables with VIP≥1 were selected and further characterised based on their fragmentation patterns ( Table 3 ). The report of a general decrease in TG concentrations and an increase in PLs are in agreement with findings describing response to PPAR-δ agonists reported by Roberts and co-workers previously [13] and with findings reported as a response to PPAR-α agonists [19] . In addition, the increases in PL concentrations may potentially provide for the required biomass during membrane development and organelle proliferation of the smooth endoplasmic reticulum (sER) and peroxisomes, effects known to be induced by PPAR-α activation [29] .…”

“…Samples were dried under nitrogen before processing for gas-chromatography–mass spectrometry (GC–MS) and liquid chromatography–mass spectrometry (LC–MS). GC–MS, and LC–MS/MS methods for lipid extraction and analysis were carried out according to methods previously described [19] , [20] .…”

Integration of metabolomics, lipidomics and clinical data using a machine learning method

“…Importantly, macrophages are elite producers of eicosanoids and other related lipid mediators during inflammation. Kupffer cells not only contribute to the production of inflammatory mediators they have also well-established connections to diet induced hepatic steatosis [28], PPAR-α activation, and non-genotoxic carcinogenesis [16]. From a metabolomics perspective, the role of macrophages in inflammatory signalling is mostly understood in the context of arachidonic acid metabolism.…”

“…Samples were dried under nitrogen before processing for GC-MS and LC-MS. GC-MS, and LC-MS/MS methods for lipid extraction and analysis were carried out according to methods previously described [16, 17]. Methods published in previous publications are also provided as supplementary for convenience (Additional file 2).…”

Using fragmentation trees and mass spectral trees for identifying unknown compounds in metabolomics

“…DD-ITe was used for structural analyses of seven vergauamide compounds in marine cyanobacteria, using a maximum MS 3 depth [38]. Non-genotoxic carcinogens in rats and lipid species were also structurally elucidated and confirmed using dd-ITe [39]. However, there are alternatives to ion-trap-based MS n studies.…”