A mathematical modelling approach to assessing the reliability of biomarkers of glutathione metabolism

https://doi.org/10.1016/j.ejps.2011.08.017Get rights and content

Abstract

One of the main pathways for the detoxification of reactive metabolites in the liver involves glutathione conjugation. Metabolic profiling studies have shown paradoxical responses in glutathione-related biochemical pathways. One of these is the increase in 5-oxoproline and ophthalmic acid concentrations with increased dosage of paracetamol. Experimental studies have thus far failed to resolve these paradoxes and the robustness of how these proposed biomarkers correlate with liver glutathione levels has been questioned. To better understand how these biomarkers behave in the glutathione system a kinetic model of this pathway was made. By using metabolic control analysis and by simulating biomarker levels under a variety of conditions, we found that 5-oxoproline and ophthalmic acid concentrations may not only depend on the glutathione but also on the methionine status of the cell. We show that neither of the two potential biomarkers are reliable on their own since they need additional information about the methionine status of the system to relate them uniquely to intracellular glutathione concentration. However, when both biomarkers are measured simultaneously a direct inference of the glutathione concentration can be made, irrespective of the methionine concentration in the system.

Introduction

Glutathione is often associated with toxicology due to its role in detoxifying electrophilic metabolites leading to the production of N-acetyl cysteinyl conjugates (mercapturates) (Meister, 1988). This is particularly true in the case of paracetamol (acetaminophen, APAP). APAP is subject to cytochrome (CYP) P450-based oxidative metabolism via CYP 2E1. This results in the formation of a reactive metabolite that is nucleophilic and at high therapeutic doses is detoxified by reacting with cellular glutathione to form glutathione conjugates. At high concentrations of paracetamol this protective mechanism depletes the system for glutathione leading to high levels of reactive metabolites which in turn can result in oxidative stress and covalent modification of cellular macromolecules (Mitchell et al., 1973).

Many factors can influence the rate of glutathione synthesis and therefore the amount of glutathione available in the cell for drug detoxification (Lu, 2009). These range from the expression level (regulation of enzyme production) to metabolism (effect of surrounding metabolism such as methionine cycle) to environmental (effect of nutrition and uptake of amino acids). Predicting how an individual’s glutathione level will react to a xenobiotic attack on the liver is currently difficult. Predictive biomarkers that indicate the level of glutathione metabolism could predict individual responses to drugs and thereby allow prediction of maximum drug dose levels, or aid in the assessment of glutathione levels in clinical trials (Mendrick and Schnackenberg, 2009).

Recently the tripeptide ophthalmic acid, a non-sulphur-containing structural analogue of glutathione, was suggested as a biomarker following APAP administration to mice (Soga et al., 2006). In that study, metabolic profiles were obtained, enabling the determination of global changes in a wide range of metabolites in serum and liver extracts of APAP-treated mice. Ophthalmic acid concentrations had increased within one hour after drug administration, remaining above normal for the next 3 h with significant decreases in glutathione concentrations in the same time period.

A number of studies have also shown increases in the concentration of 5-oxoproline following the administration of hepatic toxicants such as acetaminophen and bromobenzene, where CYP dependent bioactivation produces metabolites that require deactivation via glutathione dependent mechanisms, thereby resulting in glutathione depletion. Originally 5-oxoproline was identified in rat urine by NMR spectroscopy after dosing rats with paracetamol (Ghauri et al., 1993). Oxoprolinuria has also been observed in humans (Creer et al., 1989) following deficiency in enzymes of the glutamyl cycle, i.e. 5-oxoprolinase and glutathione synthase. A recent integrated metabonomic study into bromobenzene-induced hepatotoxicity also identified an increase in 5-oxoproline production in rats (Waters et al., 2006) and an LC-MS assay has shown an increase in 5-oxoproline in THLE-2E1 cell system with paracetamol dosing (Geenen et al., in press).

The studies above suggested a link between ophthalmic acid, 5-oxoproline and glutathione depletion. However, it was not investigated why a correlation in changes in their concentrations should exist and without a mechanistic interpretation of the biomarker’s functioning it is unknown whether any such correlation is robust. Recent publications did not show an increase in ophthalmic acid (Geenen et al., 2011) concentrations after methapyrilene toxicity in rat plasma samples. In addition measurements where 2E1 THLE-cells (SV40 large T antigen immortalised human liver epithelial cells transfected with individual cytochrome P450 enzymes (Macé et al., 1997)) were dosed with paracetamol, have not shown a significant increase in extracellular ophthalmic acid concentrations (Geenen et al., in press). Here we make use of a mathematical model of what is known about glutathione metabolism. We ask whether one should expect 5-oxoproline and ophthalmic acid to be robust biomarkers for glutathione depletion.

Section snippets

Steady state analysis of the model

For the construction of our mathematical model we adapted an existing model on glutathione metabolism (Reed et al., 2008a). The original model was simplified from 60 to 24 reactions, mostly by removing the folate pathway, and was extended to 41 reactions with relevant parts of drug detoxification metabolism (a schema of the model is shown in Fig. 1). After the adaptations we changed some of the model parameters (Vmax values and unknown kinetic parameters, a list of which is included in Appendix

Discussion

The effectiveness, toxicity and thereby maximal dosage of many drugs depend on hepatic detoxification pathways. Predictive biomarkers, which would allow assessment of an individual’s drug metabolism activity, might indirectly but dramatically enhance individualized therapies by allowing doses to be optimized and dose-related toxicity avoided.

A kinetic model of glutathione metabolism and its detoxification pathway for paracetamol was constructed and used for a tentative investigation of the

Modifying the Reed model

The model of glutathione metabolism was based on the model by Reed et al. (Reed et al., 2008b). The Reed model is available from the JWS Online website (Olivier and Snoep, 2004) and can be simulated in a web browser at any of the JWS Online servers (e.g. http://jjj.mib.ac.uk/webMathematica/Examples/run.jsp?modelName=reed). We coded the Reed model in Mathematica (http://www.wolfram.com) and the same steady state was obtained as calculated by Reed. The folate pathway, which was modelled very

Acknowledgement

We thank the BBSRC, EPSRC (BBD0190791, BBC0082191, BBF0035281, BBF0035521, BBF0035521, BBF0035361, BBG5302251, SySMO), EU-FP7 (BioSim, NucSys, EC-MOAN), ZON-MW (91206069) and other funders for systems biology support (http://www.systembiology.net/support).

References (38)

  • L. Oppenheimer et al.

    Glutathione synthetase. Purification from rat kidney and mapping of the substrate binding sites

    Journal of Biological Chemistry

    (1979)
  • M.J. Pabst et al.

    Glutathione S-transferase A

    Journal of Biological Chemistry

    (1974)
  • M. Pogribna et al.

    Homocysteine metabolism in children with Down syndrome: in vitro modulation

    American Journal of Human Genetics

    (2001)
  • J.L. Raucy et al.

    Acetaminophen activation by human liver cytochromes P450IIE1 and P450IA2

    Archives of Biochemistry and Biophysics

    (1989)
  • T. Soga et al.

    Differential metabolomics reveals ophthalmic acid as an oxidative stress biomarker indicating hepatic glutathione consumption

    Journal of Biological Chemistry

    (2006)
  • M. Soley et al.

    A rapid method for the estimation of amino acid concentration in liver tissue

    Journal of Biological and Biophysical Methods

    (1980)
  • P. Van Der Werf et al.

    5-Oxo-l-prolinase (l-pyroglutamate hydrolase). Purification and catalytic properties

    Journal of Biological Chemistry

    (1975)
  • C. Wu et al.

    Effects of ethionine on the free amino acids in the rat

    Journal of Biological Chemistry

    (1954)
  • G. Wu et al.

    Glutathione metabolism and its implications for health

    Journal of Nutrition

    (2004)
  • Cited by (19)

    • Liquid chromatographic methods combined with mass spectrometry in metabolomics

      2019, Proteomic and Metabolomic Approaches to Biomarker Discovery
    • A mechanistic model for thiol redox dynamics in the organogenesis stage rat conceptus

      2018, Reproductive Toxicology
      Citation Excerpt :

      We find a good correlation between model predictions and the evaluation dataset as well, indicating that the model is not only applicable to the particular experimental dataset used in this study, but it is generally applicable to predict GSH redox dynamics in the organogenesis stage rat embryo. Whereas previous models for GSH redox dynamics have focused on adult tissues or organs and are generally one-compartmental [6–13], we here present a multi-compartment model for GSH redox dynamics in the developing embryo. We thus needed to account for inter-compartmental transport of amino-acid precursors and relevant embryonic features, such as protein uptake via histiotrophic nutrition, volumetric growth and protein synthesis.

    • Reducing the effects of drug toxicity on glutathione metabolism

      2016, Journal of the Taiwan Institute of Chemical Engineers
      Citation Excerpt :

      In addition, numerous analytical algorithms have been developed to investigate the effects of environmental and genetic disturbances on the metabolic functions of an organism [9]; most of these algorithms are intended to identify novel diagnostic biomarkers of toxicity. The purpose of our study was to identify the target enzymes that can limit the damage of drug toxicity with the smallest regulatory intensity by analyzing the kinetic model of glutathione metabolism of Geenen et al. [7]. The basic goal was to reduce the effects of drug toxicity by regulating multiple enzymes, similar to composite prescriptions of traditional Chinese medicines [10].

    • Systematic review of ophthalmate as a novel biomarker of hepatic glutathione depletion

      2013, Clinical Nutrition
      Citation Excerpt :

      Up till now there are no studies that validate OPH as a biomarker for GSH depletion in patients with liver disease (e.g. liver resection/liver transplantation patients, PCM auto-intoxication patients, acute liver failure patients). There are also some paradoxical results in studies,4,6–8,10 which underline the necessity for a proof of principle study for the validation of OPH as a potential biomarker for GSH depletion in a human research model. In future studies the validity of OPH as a biomarker for GSH depletion and oxidative stress can probably be tested best in PCM intoxication patients.

    View all citing articles on Scopus
    View full text