Elsevier

Chemosphere

Volume 71, Issue 3, March 2008, Pages 601-609
Chemosphere

A metabolomics based approach to assessing the toxicity of the polyaromatic hydrocarbon pyrene to the earthworm Lumbricus rubellus

https://doi.org/10.1016/j.chemosphere.2007.08.056Get rights and content

Abstract

The biochemical response of the earthworm Lumbricus rubellus to pyrene exposure was assessed using 1H nuclear magnetic resonance (NMR) spectroscopy, gas chromatography mass spectrometry (GC–MS) and pattern recognition techniques. Both analytical methods enabled the establishment of reproducible metabolic profiles. NMR analysis identified a total of 32 metabolites while GC–MS identified 51. The results demonstrate that not only is pyrene toxic to L. rubellus, but that alterations in its normal metabolic profile could be observed even when individuals were exposed to concentrations of 40 mg kg−1: a pollution level that is both below the concentration previously found to significantly reduce reproduction and within the range of polycyclic aromatic hydrocarbons (PAHs) found on some contaminated sites. Pyrene was found to cause a dose dependant decrease in lactate and the concentrations of the saturated fatty acids tetradecanoic, hexadecanoic and octadecanoic acid and an increase in production of the amino acids alanine, leucine, valine, isoleucine, lysine, tyrosine and methionine. It is proposed that this indicates impaired glucose metabolism, with an associated increase in fatty acid metabolism and changes in TCA cycle intermediates. This study demonstrates the versatility of metabolomics as a tool to monitor toxicity in the environment as opposed to utilising model species studied in a laboratory setting. Since it is a non-carcinogenic PAH, we propose that the metabolic changes observed in worms may reflect the non-specific toxic effects of pyrene as a typical, non-polar organic compound.

Introduction

Polyaromatic hydrocarbons (PAHs) are hydrocarbons containing multiple-ring structures. They are mainly formed from the incomplete combustion of organic substances and so are ubiquitous environmental contaminants. Numerous members of this group have carcinogenic properties (Baek et al., 1991) and the potential to interfere with hormonal systems and immune responses (Coles et al., 1994). They are also very persistent in the environment. Consequently there are now legislative restrictions on their release in place in most countries, although PAH contamination is still a problem in many areas.

In this study the toxicity of pyrene has been assessed by metabolomics in the earthworm L. rubellus. The specific toxic effects of this compound have been studied in a wide range of species, including mussels (Okay et al., 2006), marine polychaetes (Giessing et al., 2003), springtails (Stroomberg et al., 2004), and earthworms (Brown et al., 2004). However, there are less data available regarding the underlying biochemical effects of exposure to environmentally relevant concentrations.

Metabolomics is the study of the complete set of metabolites/low molecular weight intermediates, which are context dependent, varying according to the physiology, developmental or pathological state of the cell, tissue, organ or organism (Oliver, 2002). This approach has proven to be highly sensitive for the detection of effects associated with both drugs and environmental toxins/toxicants, in that metabolic perturbations often present much earlier than pollutant induced histopathological changes (Griffin et al., 2000).

For many years a major analytical method for metabolomic studies has been nuclear magnetic resonance (NMR) spectroscopy. This has a number of advantages in that it requires minimal sample preparation and is fast and a robust technique, which allows a wide range of small molecule metabolite to be measured simultaneously. Its major disadvantage is a lack of sensitivity. For this reason many metabolomic studies also additional analytical techniques such as use gas chromatography mass spectrometry (GC–MS). This has the advantage of greatly enhanced sensitivity compared to NMR, but the disadvantages of increased sample preparation time and the fact that some large and/or very polar metabolites, such as some hormones, cannot be analyzed.

In this study we have used a combined NMR and GC–MS based approach to assess the biochemical changes induced in the metabolic profile of L. rubellus by exposure to pyrene as a model non-carcinogenic PAH. This compound is thought to work by a narcosis (membrane accumulation) based mode of action only. Hence, any changes detected in metabolic pathways in response to exposure to these compounds may indicate the metabolic consequences of non-specific ‘baseline’ narcotic toxic effects.

Section snippets

Experimental design

The experimental setup used has previously been reported in Brown et al. (2004) but is summarized briefly below. The soil used was sterilized loam and was prepared as described by Spurgeon et al. (2003). One kilogram (dry weight) of soil mix was placed dry into 1 l Kilner jars and spiked with concentrations of 0, 10, 40, 160 and 640 mg kg−1 of pyrene in equal volumes of acetone. Soils were then vented for 72 h to remove all the solvent, wetted to 60% of water holding capacity and then left to

Toxicity test results

Results from the toxicity tests have previously been presented in detail Brown et al. (2004), but are summarized here in order to support phenotypic anchoring of the metabolomic data (Paules, 2003). All worms survived in the experimental controls and there was >90% survival at the two lowest pyrene concentrations indicating that the test was robustly conducted. A significant effect of pyrene on earthworm survival (P < 0.001) was, however, found with mortality significantly (P < 0.001) higher at 640 

Conclusions

This work demonstrates the suitability of NMR and GC–MS, combined with multivariate analysis to recognise the normal biochemistry of a whole organism of relevance to ecotoxicology, as well as their metabolic responses to toxic insult. The evidence presented here confirms that pyrene is toxic to the earthworm L. rubellus. Effects were seen at concentrations of 40 mg kg−1 in this study and, although normal background levels of this compound are much lower (Maliszewska-Kordybach, 1996) contamination

Acknowledgments

This research was financially supported by the European Union (European Commission, FP6 Contract No. 003956). Julian Griffin is sponsored by the Royal Society (UK). The authors are very grateful to Dr. Mahon Maguire for help with statistical analysis, and to several anonymous reviewers whose comments greatly improved the manuscript.

References (31)

Cited by (117)

View all citing articles on Scopus
View full text