Integrating in vitro metabolomics with a 96-well high-throughput screening platform

Maurice Whelan,Mark R. Viant,Ralf J. M. Weber,Donatella Carpi,Jukka Sund,Julia M. Malinowska,Taina Palosaari,Mounir Bouhifd

doi:10.1007/s11306-021-01867-3

Maurice Whelan, Mark R. Viant + Show 6 more

Open Access

https://doi.org/10.1007/s11306-021-01867-3

Copy DOI

Abstract

IntroductionHigh-throughput screening (HTS) is emerging as an approach to support decision-making in chemical safety assessments. In parallel, in vitro metabolomics is a promising approach that can help accelerate the transition from animal models to high-throughput cell-based models in toxicity testing.ObjectiveIn this study we establish and evaluate a high-throughput metabolomics workflow that is compatible with a 96-well HTS platform employing 50,000 hepatocytes of HepaRG per well.MethodsLow biomass cell samples were extracted for metabolomics analyses using a newly established semi-automated protocol, and the intracellular metabolites were analysed using a high-resolution spectral-stitching nanoelectrospray direct infusion mass spectrometry (nESI-DIMS) method that was modified for low sample biomass.ResultsThe method was assessed with respect to sensitivity and repeatability of the entire workflow from cell culturing and sampling to measurement of the metabolic phenotype, demonstrating sufficient sensitivity (> 3000 features in hepatocyte extracts) and intra- and inter-plate repeatability for polar nESI-DIMS assays (median relative standard deviation < 30%). The assays were employed for a proof-of-principle toxicological study with a model toxicant, cadmium chloride, revealing changes in the metabolome across five sampling times in the 48-h exposure period. To allow the option for lipidomics analyses, the solvent system was extended by establishing separate extraction methods for polar metabolites and lipids.ConclusionsExperimental, analytical and informatics workflows reported here met pre-defined criteria in terms of sensitivity, repeatability and ability to detect metabolome changes induced by a toxicant and are ready for application in metabolomics-driven toxicity testing to complement HTS assays.

Highlights

High-throughput screening (HTS) is emerging as an approach to support decision-making in chemical safety assessments
The analyses presented below focus on the polar negative nanoflow electrospray ionisation (nESI)-direct infusion mass spectrometry (DIMS) assay; the final dataset used for statistical analysis demonstrated a higher technical quality than positive ionisation mode, with a feature count of 4983 and median RSD (mRSD) of intrastudy quality control (QC) samples of only 9.1%
The initially proposed monophasic solvent system (1:3:1 (v/v/v) water:methanol:chloroform) achieved the target criteria for sensitivity and repeatability when used for polar nESI-DIMS assays, issues with spray stability during lipid analysis were encountered prompting an evaluation of separate extraction methods for polar metabolites and lipids that achieved the target criteria when using 4:1 (v/v) methanol:water for polar metabolites and 2:1 (v/v) methanol:chloroform, for lipids

Summary

Introduction

High-throughput screening (HTS) is emerging as an approach to support decision-making in chemical safety assessments. The report envisioned changes to existing methods recognising that animal-based toxicity testing, usually focused on apical endpoints (e.g. histopathology), holds little relevance to humans and is unfit to test the growing number of chemicals on the market due to its low throughput. Omics technologies align well with the described shift in toxicity testing due to their ability to generate molecular mechanistic information to support decision making in regulatory applications, including chemical grouping and mode of action (MoA) prediction (Sperber et al, 2019; van Ravenzwaay et al, 2016; Viant et al, 2019). In vitro metabolomics has been demonstrated to predict organ toxicity and identify the MoA of chemicals, which could support regulatory applications and contribute towards the use of non-animal models in toxicity testing (Ramirez et al, 2018). There has been growing interest in applying in vitro metabolomics to study hepatotoxicity due to the central role of the liver in metabolism of xenobiotics (Cuykx et al, 2018a, 2018b; Mennecozzi et al, 2012; Pomponio et al, 2015; Van den Eede et al, 2015)

Objectives

Methods

Results

Conclusion